{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.0631463871386596, "eval_steps": 500, "global_step": 8000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "grad_norm": 9617.3056640625, "learning_rate": 5e-06, "loss": 675.5383, "step": 5 }, { "epoch": 0.0, "grad_norm": 8651.365234375, "learning_rate": 1e-05, "loss": 585.8184, "step": 10 }, { "epoch": 0.0, "grad_norm": 6962.1630859375, "learning_rate": 1.5e-05, "loss": 441.9639, "step": 15 }, { "epoch": 0.0, "grad_norm": 7954.0732421875, "learning_rate": 2e-05, "loss": 289.7258, "step": 20 }, { "epoch": 0.0, "grad_norm": 3123.807861328125, "learning_rate": 2.5e-05, "loss": 162.7121, "step": 25 }, { "epoch": 0.0, "grad_norm": 5402.1259765625, "learning_rate": 3e-05, "loss": 116.8305, "step": 30 }, { "epoch": 0.0, "grad_norm": 1373.7694091796875, "learning_rate": 3.5e-05, "loss": 79.7264, "step": 35 }, { "epoch": 0.0, "grad_norm": 1104.11767578125, "learning_rate": 4e-05, "loss": 35.2114, "step": 40 }, { "epoch": 0.0, "grad_norm": 189.30984497070312, "learning_rate": 4.5e-05, "loss": 18.3442, "step": 45 }, { "epoch": 0.0, "grad_norm": 53.9228401184082, "learning_rate": 5e-05, "loss": 10.5773, "step": 50 }, { "epoch": 0.0, "grad_norm": 17.824100494384766, "learning_rate": 5.500000000000001e-05, "loss": 8.428, "step": 55 }, { "epoch": 0.0, "grad_norm": 21.479326248168945, "learning_rate": 6e-05, "loss": 7.9351, "step": 60 }, { "epoch": 0.0, "grad_norm": 10.783280372619629, "learning_rate": 6.500000000000001e-05, "loss": 7.3717, "step": 65 }, { "epoch": 0.0, "grad_norm": 5.471942901611328, "learning_rate": 7e-05, "loss": 7.1058, "step": 70 }, { "epoch": 0.0, "grad_norm": 10.43978500366211, "learning_rate": 7.500000000000001e-05, "loss": 6.8687, "step": 75 }, { "epoch": 0.0, "grad_norm": 11.331214904785156, "learning_rate": 8e-05, "loss": 6.6485, "step": 80 }, { "epoch": 0.0, "grad_norm": 8.001110076904297, "learning_rate": 8.5e-05, "loss": 6.4427, "step": 85 }, { "epoch": 0.0, "grad_norm": 5.011355400085449, "learning_rate": 9e-05, "loss": 6.154, "step": 90 }, { "epoch": 0.0, "grad_norm": 9.839451789855957, "learning_rate": 9.5e-05, "loss": 6.1568, "step": 95 }, { "epoch": 0.0, "grad_norm": 8.513458251953125, "learning_rate": 0.0001, "loss": 5.9143, "step": 100 }, { "epoch": 0.0, "grad_norm": 5.411273002624512, "learning_rate": 9.999999961506464e-05, "loss": 5.9149, "step": 105 }, { "epoch": 0.0, "grad_norm": 6.151527404785156, "learning_rate": 9.999999846025851e-05, "loss": 5.6177, "step": 110 }, { "epoch": 0.0, "grad_norm": 4.462268829345703, "learning_rate": 9.999999653558166e-05, "loss": 5.4205, "step": 115 }, { "epoch": 0.0, "grad_norm": 5.890280246734619, "learning_rate": 9.999999384103412e-05, "loss": 5.3409, "step": 120 }, { "epoch": 0.0, "grad_norm": 18.63995933532715, "learning_rate": 9.999999037661592e-05, "loss": 5.1337, "step": 125 }, { "epoch": 0.0, "grad_norm": 5.008910655975342, "learning_rate": 9.999998614232713e-05, "loss": 4.9855, "step": 130 }, { "epoch": 0.0, "grad_norm": 9.032185554504395, "learning_rate": 9.99999811381678e-05, "loss": 4.9934, "step": 135 }, { "epoch": 0.0, "grad_norm": 3.891791343688965, "learning_rate": 9.9999975364138e-05, "loss": 4.913, "step": 140 }, { "epoch": 0.0, "grad_norm": 3.245584011077881, "learning_rate": 9.999996882023784e-05, "loss": 4.6751, "step": 145 }, { "epoch": 0.0, "grad_norm": 3.871345281600952, "learning_rate": 9.99999615064674e-05, "loss": 4.5819, "step": 150 }, { "epoch": 0.0, "grad_norm": 3.6179027557373047, "learning_rate": 9.99999534228268e-05, "loss": 4.5478, "step": 155 }, { "epoch": 0.0, "grad_norm": 6.745605945587158, "learning_rate": 9.999994456931619e-05, "loss": 4.5344, "step": 160 }, { "epoch": 0.0, "grad_norm": 6.126508712768555, "learning_rate": 9.999993494593566e-05, "loss": 4.3437, "step": 165 }, { "epoch": 0.0, "grad_norm": 4.00099515914917, "learning_rate": 9.999992455268539e-05, "loss": 4.4348, "step": 170 }, { "epoch": 0.0, "grad_norm": 3.859551429748535, "learning_rate": 9.999991338956554e-05, "loss": 4.208, "step": 175 }, { "epoch": 0.0, "grad_norm": 4.496798515319824, "learning_rate": 9.999990145657626e-05, "loss": 4.1429, "step": 180 }, { "epoch": 0.0, "grad_norm": 3.5526297092437744, "learning_rate": 9.999988875371776e-05, "loss": 4.1706, "step": 185 }, { "epoch": 0.0, "grad_norm": 3.6115028858184814, "learning_rate": 9.999987528099021e-05, "loss": 4.0555, "step": 190 }, { "epoch": 0.0, "grad_norm": 3.514538288116455, "learning_rate": 9.999986103839384e-05, "loss": 4.0721, "step": 195 }, { "epoch": 0.0, "grad_norm": 41.726566314697266, "learning_rate": 9.999984602592884e-05, "loss": 4.0301, "step": 200 }, { "epoch": 0.0, "grad_norm": 15.108702659606934, "learning_rate": 9.999983024359549e-05, "loss": 4.0353, "step": 205 }, { "epoch": 0.0, "grad_norm": 2.9150753021240234, "learning_rate": 9.9999813691394e-05, "loss": 3.9192, "step": 210 }, { "epoch": 0.0, "grad_norm": 6.2293314933776855, "learning_rate": 9.999979636932461e-05, "loss": 3.8793, "step": 215 }, { "epoch": 0.0, "grad_norm": 2.9548492431640625, "learning_rate": 9.999977827738762e-05, "loss": 3.8549, "step": 220 }, { "epoch": 0.0, "grad_norm": 151.67041015625, "learning_rate": 9.999975941558328e-05, "loss": 3.9205, "step": 225 }, { "epoch": 0.0, "grad_norm": 3.0722227096557617, "learning_rate": 9.999973978391191e-05, "loss": 3.8946, "step": 230 }, { "epoch": 0.0, "grad_norm": 2.9939091205596924, "learning_rate": 9.99997193823738e-05, "loss": 3.7565, "step": 235 }, { "epoch": 0.0, "grad_norm": 3.441255569458008, "learning_rate": 9.999969821096924e-05, "loss": 3.6388, "step": 240 }, { "epoch": 0.0, "grad_norm": 30.157629013061523, "learning_rate": 9.99996762696986e-05, "loss": 3.7767, "step": 245 }, { "epoch": 0.0, "grad_norm": 2.106005907058716, "learning_rate": 9.999965355856218e-05, "loss": 3.7458, "step": 250 }, { "epoch": 0.0, "grad_norm": 5.746977806091309, "learning_rate": 9.999963007756035e-05, "loss": 3.6809, "step": 255 }, { "epoch": 0.0, "grad_norm": 3.0916407108306885, "learning_rate": 9.999960582669346e-05, "loss": 3.5982, "step": 260 }, { "epoch": 0.0, "grad_norm": 1.9926202297210693, "learning_rate": 9.99995808059619e-05, "loss": 3.6992, "step": 265 }, { "epoch": 0.0, "grad_norm": 12.389793395996094, "learning_rate": 9.999955501536603e-05, "loss": 3.6328, "step": 270 }, { "epoch": 0.0, "grad_norm": 5.994906425476074, "learning_rate": 9.999952845490626e-05, "loss": 3.6528, "step": 275 }, { "epoch": 0.0, "grad_norm": 2.7915570735931396, "learning_rate": 9.999950112458302e-05, "loss": 3.5915, "step": 280 }, { "epoch": 0.0, "grad_norm": 3.208523988723755, "learning_rate": 9.99994730243967e-05, "loss": 3.5161, "step": 285 }, { "epoch": 0.0, "grad_norm": 18.24662208557129, "learning_rate": 9.999944415434774e-05, "loss": 3.5449, "step": 290 }, { "epoch": 0.0, "grad_norm": 1.67327082157135, "learning_rate": 9.99994145144366e-05, "loss": 3.4749, "step": 295 }, { "epoch": 0.0, "grad_norm": 2.082451820373535, "learning_rate": 9.999938410466371e-05, "loss": 3.5448, "step": 300 }, { "epoch": 0.0, "grad_norm": 2.4315760135650635, "learning_rate": 9.999935292502958e-05, "loss": 3.6121, "step": 305 }, { "epoch": 0.0, "grad_norm": 2.2345972061157227, "learning_rate": 9.999932097553463e-05, "loss": 3.5473, "step": 310 }, { "epoch": 0.0, "grad_norm": 19.251840591430664, "learning_rate": 9.999928825617941e-05, "loss": 3.5329, "step": 315 }, { "epoch": 0.0, "grad_norm": 3.046705722808838, "learning_rate": 9.999925476696439e-05, "loss": 3.5839, "step": 320 }, { "epoch": 0.0, "grad_norm": 4.400516033172607, "learning_rate": 9.99992205078901e-05, "loss": 3.381, "step": 325 }, { "epoch": 0.0, "grad_norm": 26.005325317382812, "learning_rate": 9.999918547895705e-05, "loss": 3.4544, "step": 330 }, { "epoch": 0.0, "grad_norm": 24.644346237182617, "learning_rate": 9.999914968016579e-05, "loss": 3.4503, "step": 335 }, { "epoch": 0.0, "grad_norm": 2.57073712348938, "learning_rate": 9.999911311151689e-05, "loss": 3.3904, "step": 340 }, { "epoch": 0.0, "grad_norm": 1.5260676145553589, "learning_rate": 9.999907577301089e-05, "loss": 3.4985, "step": 345 }, { "epoch": 0.0, "grad_norm": 2.627871513366699, "learning_rate": 9.999903766464837e-05, "loss": 3.4745, "step": 350 }, { "epoch": 0.0, "grad_norm": 4.024744510650635, "learning_rate": 9.99989987864299e-05, "loss": 3.4432, "step": 355 }, { "epoch": 0.0, "grad_norm": 5.373614311218262, "learning_rate": 9.999895913835613e-05, "loss": 3.4204, "step": 360 }, { "epoch": 0.0, "grad_norm": 4.983832836151123, "learning_rate": 9.999891872042762e-05, "loss": 3.4168, "step": 365 }, { "epoch": 0.0, "grad_norm": 2.366701364517212, "learning_rate": 9.9998877532645e-05, "loss": 3.3318, "step": 370 }, { "epoch": 0.0, "grad_norm": 3.2913784980773926, "learning_rate": 9.999883557500892e-05, "loss": 3.3015, "step": 375 }, { "epoch": 0.0, "grad_norm": 2.2424943447113037, "learning_rate": 9.999879284752003e-05, "loss": 3.2775, "step": 380 }, { "epoch": 0.0, "grad_norm": 1.5550882816314697, "learning_rate": 9.999874935017897e-05, "loss": 3.2109, "step": 385 }, { "epoch": 0.0, "grad_norm": 2.9295287132263184, "learning_rate": 9.999870508298641e-05, "loss": 3.4355, "step": 390 }, { "epoch": 0.0, "grad_norm": 2.3778364658355713, "learning_rate": 9.999866004594303e-05, "loss": 3.4521, "step": 395 }, { "epoch": 0.0, "grad_norm": 7.854914665222168, "learning_rate": 9.999861423904954e-05, "loss": 3.3055, "step": 400 }, { "epoch": 0.0, "grad_norm": 9.703304290771484, "learning_rate": 9.999856766230665e-05, "loss": 3.2113, "step": 405 }, { "epoch": 0.0, "grad_norm": 4.191004753112793, "learning_rate": 9.999852031571506e-05, "loss": 3.3005, "step": 410 }, { "epoch": 0.0, "grad_norm": 3.380415678024292, "learning_rate": 9.999847219927549e-05, "loss": 3.2666, "step": 415 }, { "epoch": 0.0, "grad_norm": 1.424092173576355, "learning_rate": 9.999842331298872e-05, "loss": 3.3328, "step": 420 }, { "epoch": 0.0, "grad_norm": 2.5037591457366943, "learning_rate": 9.999837365685544e-05, "loss": 3.2174, "step": 425 }, { "epoch": 0.0, "grad_norm": 2.168788433074951, "learning_rate": 9.999832323087649e-05, "loss": 3.4091, "step": 430 }, { "epoch": 0.0, "grad_norm": 2.8762829303741455, "learning_rate": 9.999827203505259e-05, "loss": 3.3185, "step": 435 }, { "epoch": 0.0, "grad_norm": 2.046619176864624, "learning_rate": 9.999822006938455e-05, "loss": 3.3322, "step": 440 }, { "epoch": 0.0, "grad_norm": 1.4038219451904297, "learning_rate": 9.999816733387316e-05, "loss": 3.2711, "step": 445 }, { "epoch": 0.0, "grad_norm": 1.5367592573165894, "learning_rate": 9.999811382851924e-05, "loss": 3.2171, "step": 450 }, { "epoch": 0.0, "grad_norm": 3.4895827770233154, "learning_rate": 9.999805955332362e-05, "loss": 3.2953, "step": 455 }, { "epoch": 0.0, "grad_norm": 1.4623072147369385, "learning_rate": 9.999800450828711e-05, "loss": 3.1775, "step": 460 }, { "epoch": 0.0, "grad_norm": 1.6363601684570312, "learning_rate": 9.99979486934106e-05, "loss": 3.2731, "step": 465 }, { "epoch": 0.0, "grad_norm": 3.7503108978271484, "learning_rate": 9.99978921086949e-05, "loss": 3.1347, "step": 470 }, { "epoch": 0.0, "grad_norm": 1.5456328392028809, "learning_rate": 9.999783475414093e-05, "loss": 3.219, "step": 475 }, { "epoch": 0.0, "grad_norm": 1.1970163583755493, "learning_rate": 9.999777662974952e-05, "loss": 3.1819, "step": 480 }, { "epoch": 0.0, "grad_norm": 1.1249005794525146, "learning_rate": 9.999771773552162e-05, "loss": 3.1699, "step": 485 }, { "epoch": 0.0, "grad_norm": 4.346677303314209, "learning_rate": 9.999765807145811e-05, "loss": 3.2317, "step": 490 }, { "epoch": 0.0, "grad_norm": 1.474409580230713, "learning_rate": 9.99975976375599e-05, "loss": 3.1972, "step": 495 }, { "epoch": 0.0, "grad_norm": 2.134998083114624, "learning_rate": 9.999753643382793e-05, "loss": 3.1837, "step": 500 }, { "epoch": 0.0, "grad_norm": 2.396498918533325, "learning_rate": 9.999747446026315e-05, "loss": 3.1851, "step": 505 }, { "epoch": 0.0, "grad_norm": 1.664106011390686, "learning_rate": 9.999741171686649e-05, "loss": 3.0756, "step": 510 }, { "epoch": 0.0, "grad_norm": 14.095783233642578, "learning_rate": 9.999734820363895e-05, "loss": 3.1794, "step": 515 }, { "epoch": 0.0, "grad_norm": 2.191141128540039, "learning_rate": 9.99972839205815e-05, "loss": 3.298, "step": 520 }, { "epoch": 0.0, "grad_norm": 1.4646538496017456, "learning_rate": 9.999721886769511e-05, "loss": 3.1778, "step": 525 }, { "epoch": 0.0, "grad_norm": 1.7615983486175537, "learning_rate": 9.999715304498078e-05, "loss": 3.0307, "step": 530 }, { "epoch": 0.0, "grad_norm": 5.03188943862915, "learning_rate": 9.999708645243955e-05, "loss": 3.1307, "step": 535 }, { "epoch": 0.0, "grad_norm": 1.3999274969100952, "learning_rate": 9.999701909007242e-05, "loss": 3.107, "step": 540 }, { "epoch": 0.0, "grad_norm": 3.252723455429077, "learning_rate": 9.999695095788044e-05, "loss": 3.1973, "step": 545 }, { "epoch": 0.0, "grad_norm": 7.995652198791504, "learning_rate": 9.999688205586468e-05, "loss": 3.0912, "step": 550 }, { "epoch": 0.0, "grad_norm": 26.716060638427734, "learning_rate": 9.999681238402615e-05, "loss": 3.2052, "step": 555 }, { "epoch": 0.0, "grad_norm": 1.7161030769348145, "learning_rate": 9.999674194236597e-05, "loss": 3.1535, "step": 560 }, { "epoch": 0.0, "grad_norm": 2.3022992610931396, "learning_rate": 9.999667073088519e-05, "loss": 3.0991, "step": 565 }, { "epoch": 0.0, "grad_norm": 2.029336452484131, "learning_rate": 9.999659874958493e-05, "loss": 3.0884, "step": 570 }, { "epoch": 0.0, "grad_norm": 2.5739989280700684, "learning_rate": 9.99965259984663e-05, "loss": 3.0581, "step": 575 }, { "epoch": 0.0, "grad_norm": 1.4700506925582886, "learning_rate": 9.999645247753039e-05, "loss": 3.1031, "step": 580 }, { "epoch": 0.0, "grad_norm": 4.734466075897217, "learning_rate": 9.999637818677836e-05, "loss": 3.1233, "step": 585 }, { "epoch": 0.0, "grad_norm": 100.78364562988281, "learning_rate": 9.999630312621136e-05, "loss": 3.1046, "step": 590 }, { "epoch": 0.0, "grad_norm": 3.624537229537964, "learning_rate": 9.999622729583051e-05, "loss": 3.266, "step": 595 }, { "epoch": 0.0, "grad_norm": 377.94500732421875, "learning_rate": 9.999615069563701e-05, "loss": 3.4809, "step": 600 }, { "epoch": 0.0, "grad_norm": 2.086933135986328, "learning_rate": 9.999607332563205e-05, "loss": 3.1383, "step": 605 }, { "epoch": 0.0, "grad_norm": 1.6288503408432007, "learning_rate": 9.999599518581676e-05, "loss": 3.0628, "step": 610 }, { "epoch": 0.0, "grad_norm": 7.008534908294678, "learning_rate": 9.999591627619242e-05, "loss": 3.132, "step": 615 }, { "epoch": 0.0, "grad_norm": 1.8186450004577637, "learning_rate": 9.999583659676019e-05, "loss": 3.0545, "step": 620 }, { "epoch": 0.0, "grad_norm": 1.6767669916152954, "learning_rate": 9.999575614752133e-05, "loss": 3.0289, "step": 625 }, { "epoch": 0.0, "grad_norm": 2.8385417461395264, "learning_rate": 9.999567492847706e-05, "loss": 3.0825, "step": 630 }, { "epoch": 0.01, "grad_norm": 1.6340068578720093, "learning_rate": 9.999559293962864e-05, "loss": 3.039, "step": 635 }, { "epoch": 0.01, "grad_norm": 28.434417724609375, "learning_rate": 9.99955101809773e-05, "loss": 3.0498, "step": 640 }, { "epoch": 0.01, "grad_norm": 1.2716106176376343, "learning_rate": 9.999542665252437e-05, "loss": 3.0833, "step": 645 }, { "epoch": 0.01, "grad_norm": 1.425293207168579, "learning_rate": 9.999534235427109e-05, "loss": 3.037, "step": 650 }, { "epoch": 0.01, "grad_norm": 2.7019450664520264, "learning_rate": 9.999525728621879e-05, "loss": 3.0662, "step": 655 }, { "epoch": 0.01, "grad_norm": 11.211701393127441, "learning_rate": 9.999517144836876e-05, "loss": 3.1467, "step": 660 }, { "epoch": 0.01, "grad_norm": 1.2351415157318115, "learning_rate": 9.999508484072233e-05, "loss": 3.0616, "step": 665 }, { "epoch": 0.01, "grad_norm": 12.54974365234375, "learning_rate": 9.999499746328083e-05, "loss": 3.1077, "step": 670 }, { "epoch": 0.01, "grad_norm": 1.3160040378570557, "learning_rate": 9.99949093160456e-05, "loss": 3.0997, "step": 675 }, { "epoch": 0.01, "grad_norm": 1.603757381439209, "learning_rate": 9.9994820399018e-05, "loss": 3.1483, "step": 680 }, { "epoch": 0.01, "grad_norm": 1.2601557970046997, "learning_rate": 9.999473071219941e-05, "loss": 3.019, "step": 685 }, { "epoch": 0.01, "grad_norm": 2.4634552001953125, "learning_rate": 9.999464025559121e-05, "loss": 2.9965, "step": 690 }, { "epoch": 0.01, "grad_norm": 1.965299129486084, "learning_rate": 9.999454902919478e-05, "loss": 3.0272, "step": 695 }, { "epoch": 0.01, "grad_norm": 3.274756908416748, "learning_rate": 9.999445703301154e-05, "loss": 3.0977, "step": 700 }, { "epoch": 0.01, "grad_norm": 1.9171090126037598, "learning_rate": 9.999436426704287e-05, "loss": 3.0596, "step": 705 }, { "epoch": 0.01, "grad_norm": 3.7423007488250732, "learning_rate": 9.999427073129026e-05, "loss": 3.1134, "step": 710 }, { "epoch": 0.01, "grad_norm": 1.6005221605300903, "learning_rate": 9.999417642575509e-05, "loss": 3.0238, "step": 715 }, { "epoch": 0.01, "grad_norm": 1.7754089832305908, "learning_rate": 9.999408135043886e-05, "loss": 3.0121, "step": 720 }, { "epoch": 0.01, "grad_norm": 2.796038866043091, "learning_rate": 9.999398550534298e-05, "loss": 3.0424, "step": 725 }, { "epoch": 0.01, "grad_norm": 2.449819326400757, "learning_rate": 9.999388889046899e-05, "loss": 3.1031, "step": 730 }, { "epoch": 0.01, "grad_norm": 1.903029441833496, "learning_rate": 9.999379150581833e-05, "loss": 2.9926, "step": 735 }, { "epoch": 0.01, "grad_norm": 1.4327459335327148, "learning_rate": 9.999369335139251e-05, "loss": 3.0117, "step": 740 }, { "epoch": 0.01, "grad_norm": 2.1567108631134033, "learning_rate": 9.999359442719306e-05, "loss": 3.004, "step": 745 }, { "epoch": 0.01, "grad_norm": 1.5228947401046753, "learning_rate": 9.999349473322149e-05, "loss": 3.004, "step": 750 }, { "epoch": 0.01, "grad_norm": 2.5921003818511963, "learning_rate": 9.999339426947932e-05, "loss": 3.0169, "step": 755 }, { "epoch": 0.01, "grad_norm": 2.8633832931518555, "learning_rate": 9.999329303596812e-05, "loss": 2.9759, "step": 760 }, { "epoch": 0.01, "grad_norm": 60.35059356689453, "learning_rate": 9.999319103268945e-05, "loss": 3.0306, "step": 765 }, { "epoch": 0.01, "grad_norm": 1.4301698207855225, "learning_rate": 9.999308825964485e-05, "loss": 3.0069, "step": 770 }, { "epoch": 0.01, "grad_norm": 3.65624737739563, "learning_rate": 9.999298471683593e-05, "loss": 3.024, "step": 775 }, { "epoch": 0.01, "grad_norm": 12.166756629943848, "learning_rate": 9.999288040426429e-05, "loss": 3.0152, "step": 780 }, { "epoch": 0.01, "grad_norm": 2.0820701122283936, "learning_rate": 9.999277532193151e-05, "loss": 3.0538, "step": 785 }, { "epoch": 0.01, "grad_norm": 17.54198455810547, "learning_rate": 9.999266946983923e-05, "loss": 3.006, "step": 790 }, { "epoch": 0.01, "grad_norm": 1.2765568494796753, "learning_rate": 9.999256284798906e-05, "loss": 2.9247, "step": 795 }, { "epoch": 0.01, "grad_norm": 3.1964027881622314, "learning_rate": 9.999245545638267e-05, "loss": 3.04, "step": 800 }, { "epoch": 0.01, "grad_norm": 1.7309542894363403, "learning_rate": 9.999234729502167e-05, "loss": 3.0914, "step": 805 }, { "epoch": 0.01, "grad_norm": 2.3562445640563965, "learning_rate": 9.999223836390777e-05, "loss": 2.9635, "step": 810 }, { "epoch": 0.01, "grad_norm": 1.2265539169311523, "learning_rate": 9.999212866304262e-05, "loss": 2.9538, "step": 815 }, { "epoch": 0.01, "grad_norm": 1.9237613677978516, "learning_rate": 9.999201819242794e-05, "loss": 2.9183, "step": 820 }, { "epoch": 0.01, "grad_norm": 1.4826682806015015, "learning_rate": 9.99919069520654e-05, "loss": 3.0891, "step": 825 }, { "epoch": 0.01, "grad_norm": 40.569236755371094, "learning_rate": 9.99917949419567e-05, "loss": 3.0115, "step": 830 }, { "epoch": 0.01, "grad_norm": 1.2582329511642456, "learning_rate": 9.999168216210362e-05, "loss": 2.9485, "step": 835 }, { "epoch": 0.01, "grad_norm": 2.5545523166656494, "learning_rate": 9.999156861250784e-05, "loss": 2.871, "step": 840 }, { "epoch": 0.01, "grad_norm": 3.8551647663116455, "learning_rate": 9.999145429317114e-05, "loss": 3.0069, "step": 845 }, { "epoch": 0.01, "grad_norm": 1.3857111930847168, "learning_rate": 9.999133920409527e-05, "loss": 2.9259, "step": 850 }, { "epoch": 0.01, "grad_norm": 2.373899459838867, "learning_rate": 9.999122334528199e-05, "loss": 3.1004, "step": 855 }, { "epoch": 0.01, "grad_norm": 1.1470832824707031, "learning_rate": 9.999110671673313e-05, "loss": 2.9228, "step": 860 }, { "epoch": 0.01, "grad_norm": 3.8054158687591553, "learning_rate": 9.999098931845044e-05, "loss": 2.8785, "step": 865 }, { "epoch": 0.01, "grad_norm": 1.7046176195144653, "learning_rate": 9.999087115043573e-05, "loss": 2.9652, "step": 870 }, { "epoch": 0.01, "grad_norm": 5.047961235046387, "learning_rate": 9.999075221269083e-05, "loss": 2.9891, "step": 875 }, { "epoch": 0.01, "grad_norm": 1.1203137636184692, "learning_rate": 9.999063250521759e-05, "loss": 2.9878, "step": 880 }, { "epoch": 0.01, "grad_norm": 1.7854154109954834, "learning_rate": 9.999051202801781e-05, "loss": 2.9036, "step": 885 }, { "epoch": 0.01, "grad_norm": 1.1594346761703491, "learning_rate": 9.999039078109341e-05, "loss": 2.9004, "step": 890 }, { "epoch": 0.01, "grad_norm": 18.54465675354004, "learning_rate": 9.999026876444617e-05, "loss": 2.8969, "step": 895 }, { "epoch": 0.01, "grad_norm": 1.3541970252990723, "learning_rate": 9.999014597807805e-05, "loss": 2.9552, "step": 900 }, { "epoch": 0.01, "grad_norm": 4.923300266265869, "learning_rate": 9.99900224219909e-05, "loss": 2.95, "step": 905 }, { "epoch": 0.01, "grad_norm": 10.090089797973633, "learning_rate": 9.998989809618662e-05, "loss": 2.9389, "step": 910 }, { "epoch": 0.01, "grad_norm": 2.6130008697509766, "learning_rate": 9.998977300066715e-05, "loss": 2.9633, "step": 915 }, { "epoch": 0.01, "grad_norm": 4.81696891784668, "learning_rate": 9.998964713543437e-05, "loss": 3.0181, "step": 920 }, { "epoch": 0.01, "grad_norm": 1.2481880187988281, "learning_rate": 9.998952050049026e-05, "loss": 2.8994, "step": 925 }, { "epoch": 0.01, "grad_norm": 0.9523406624794006, "learning_rate": 9.998939309583677e-05, "loss": 2.8564, "step": 930 }, { "epoch": 0.01, "grad_norm": 2.0296237468719482, "learning_rate": 9.998926492147585e-05, "loss": 2.938, "step": 935 }, { "epoch": 0.01, "grad_norm": 1.4497069120407104, "learning_rate": 9.998913597740945e-05, "loss": 2.9111, "step": 940 }, { "epoch": 0.01, "grad_norm": 1.2510639429092407, "learning_rate": 9.99890062636396e-05, "loss": 2.9186, "step": 945 }, { "epoch": 0.01, "grad_norm": 2.5958635807037354, "learning_rate": 9.998887578016827e-05, "loss": 2.9926, "step": 950 }, { "epoch": 0.01, "grad_norm": 1.9045034646987915, "learning_rate": 9.998874452699747e-05, "loss": 2.9348, "step": 955 }, { "epoch": 0.01, "grad_norm": 2.2150914669036865, "learning_rate": 9.998861250412922e-05, "loss": 2.8073, "step": 960 }, { "epoch": 0.01, "grad_norm": 2.594472646713257, "learning_rate": 9.998847971156556e-05, "loss": 3.0636, "step": 965 }, { "epoch": 0.01, "grad_norm": 0.9636615514755249, "learning_rate": 9.998834614930855e-05, "loss": 2.9364, "step": 970 }, { "epoch": 0.01, "grad_norm": 0.9943350553512573, "learning_rate": 9.99882118173602e-05, "loss": 2.8584, "step": 975 }, { "epoch": 0.01, "grad_norm": 71.1482162475586, "learning_rate": 9.998807671572263e-05, "loss": 2.9462, "step": 980 }, { "epoch": 0.01, "grad_norm": 4.120433807373047, "learning_rate": 9.998794084439789e-05, "loss": 2.9423, "step": 985 }, { "epoch": 0.01, "grad_norm": 2.7740607261657715, "learning_rate": 9.998780420338809e-05, "loss": 3.0078, "step": 990 }, { "epoch": 0.01, "grad_norm": 1.5119786262512207, "learning_rate": 9.998766679269531e-05, "loss": 2.8228, "step": 995 }, { "epoch": 0.01, "grad_norm": 1.5010024309158325, "learning_rate": 9.998752861232168e-05, "loss": 2.7896, "step": 1000 }, { "epoch": 0.01, "grad_norm": 3.684274435043335, "learning_rate": 9.998738966226933e-05, "loss": 2.8597, "step": 1005 }, { "epoch": 0.01, "grad_norm": 1.2949076890945435, "learning_rate": 9.998724994254041e-05, "loss": 2.9009, "step": 1010 }, { "epoch": 0.01, "grad_norm": 2.285193681716919, "learning_rate": 9.998710945313705e-05, "loss": 2.9897, "step": 1015 }, { "epoch": 0.01, "grad_norm": 1.071364402770996, "learning_rate": 9.998696819406141e-05, "loss": 2.8909, "step": 1020 }, { "epoch": 0.01, "grad_norm": 1.1059892177581787, "learning_rate": 9.99868261653157e-05, "loss": 3.0065, "step": 1025 }, { "epoch": 0.01, "grad_norm": 31.39375114440918, "learning_rate": 9.998668336690206e-05, "loss": 2.8821, "step": 1030 }, { "epoch": 0.01, "grad_norm": 2.0363919734954834, "learning_rate": 9.998653979882271e-05, "loss": 2.8713, "step": 1035 }, { "epoch": 0.01, "grad_norm": 3.6545305252075195, "learning_rate": 9.998639546107989e-05, "loss": 2.8475, "step": 1040 }, { "epoch": 0.01, "grad_norm": 5.9844279289245605, "learning_rate": 9.998625035367577e-05, "loss": 2.8983, "step": 1045 }, { "epoch": 0.01, "grad_norm": 1.6407831907272339, "learning_rate": 9.998610447661261e-05, "loss": 3.0047, "step": 1050 }, { "epoch": 0.01, "grad_norm": 0.9443334937095642, "learning_rate": 9.998595782989266e-05, "loss": 2.8766, "step": 1055 }, { "epoch": 0.01, "grad_norm": 2.049736261367798, "learning_rate": 9.998581041351818e-05, "loss": 2.868, "step": 1060 }, { "epoch": 0.01, "grad_norm": 1.8569691181182861, "learning_rate": 9.998566222749142e-05, "loss": 2.8829, "step": 1065 }, { "epoch": 0.01, "grad_norm": 8.340147018432617, "learning_rate": 9.998551327181468e-05, "loss": 2.926, "step": 1070 }, { "epoch": 0.01, "grad_norm": 6.312691688537598, "learning_rate": 9.998536354649026e-05, "loss": 3.0213, "step": 1075 }, { "epoch": 0.01, "grad_norm": 2.7132656574249268, "learning_rate": 9.998521305152045e-05, "loss": 2.8204, "step": 1080 }, { "epoch": 0.01, "grad_norm": 1.4575427770614624, "learning_rate": 9.998506178690755e-05, "loss": 3.1033, "step": 1085 }, { "epoch": 0.01, "grad_norm": 1.8095273971557617, "learning_rate": 9.998490975265393e-05, "loss": 2.9315, "step": 1090 }, { "epoch": 0.01, "grad_norm": 8.884051322937012, "learning_rate": 9.99847569487619e-05, "loss": 2.9636, "step": 1095 }, { "epoch": 0.01, "grad_norm": 1.4851592779159546, "learning_rate": 9.998460337523383e-05, "loss": 2.7969, "step": 1100 }, { "epoch": 0.01, "grad_norm": 2.109487771987915, "learning_rate": 9.998444903207207e-05, "loss": 2.8907, "step": 1105 }, { "epoch": 0.01, "grad_norm": 1.883592128753662, "learning_rate": 9.998429391927902e-05, "loss": 2.9061, "step": 1110 }, { "epoch": 0.01, "grad_norm": 2.0653481483459473, "learning_rate": 9.998413803685705e-05, "loss": 2.8578, "step": 1115 }, { "epoch": 0.01, "grad_norm": 1.156416893005371, "learning_rate": 9.998398138480855e-05, "loss": 2.8327, "step": 1120 }, { "epoch": 0.01, "grad_norm": 1.2373918294906616, "learning_rate": 9.998382396313597e-05, "loss": 2.7904, "step": 1125 }, { "epoch": 0.01, "grad_norm": 1.953126311302185, "learning_rate": 9.998366577184168e-05, "loss": 2.8048, "step": 1130 }, { "epoch": 0.01, "grad_norm": 3.4578940868377686, "learning_rate": 9.998350681092817e-05, "loss": 2.8582, "step": 1135 }, { "epoch": 0.01, "grad_norm": 1.1795789003372192, "learning_rate": 9.998334708039785e-05, "loss": 2.8504, "step": 1140 }, { "epoch": 0.01, "grad_norm": 1.9085960388183594, "learning_rate": 9.998318658025318e-05, "loss": 2.7824, "step": 1145 }, { "epoch": 0.01, "grad_norm": 1.5370389223098755, "learning_rate": 9.998302531049667e-05, "loss": 2.8234, "step": 1150 }, { "epoch": 0.01, "grad_norm": 0.9367464184761047, "learning_rate": 9.998286327113076e-05, "loss": 2.9105, "step": 1155 }, { "epoch": 0.01, "grad_norm": 1.4136650562286377, "learning_rate": 9.998270046215797e-05, "loss": 2.9559, "step": 1160 }, { "epoch": 0.01, "grad_norm": 3.950904130935669, "learning_rate": 9.998253688358078e-05, "loss": 2.9057, "step": 1165 }, { "epoch": 0.01, "grad_norm": 1.9507569074630737, "learning_rate": 9.998237253540174e-05, "loss": 2.8938, "step": 1170 }, { "epoch": 0.01, "grad_norm": 3.561965227127075, "learning_rate": 9.998220741762338e-05, "loss": 2.8604, "step": 1175 }, { "epoch": 0.01, "grad_norm": 7.1781134605407715, "learning_rate": 9.998204153024821e-05, "loss": 2.8854, "step": 1180 }, { "epoch": 0.01, "grad_norm": 0.9091061353683472, "learning_rate": 9.998187487327882e-05, "loss": 2.7619, "step": 1185 }, { "epoch": 0.01, "grad_norm": 3.555330514907837, "learning_rate": 9.998170744671775e-05, "loss": 2.89, "step": 1190 }, { "epoch": 0.01, "grad_norm": 1.55420982837677, "learning_rate": 9.998153925056758e-05, "loss": 2.8087, "step": 1195 }, { "epoch": 0.01, "grad_norm": 39.49836349487305, "learning_rate": 9.998137028483092e-05, "loss": 2.9578, "step": 1200 }, { "epoch": 0.01, "grad_norm": 2.1230616569519043, "learning_rate": 9.998120054951035e-05, "loss": 2.8534, "step": 1205 }, { "epoch": 0.01, "grad_norm": 310.4348449707031, "learning_rate": 9.99810300446085e-05, "loss": 3.1607, "step": 1210 }, { "epoch": 0.01, "grad_norm": 1.8449273109436035, "learning_rate": 9.9980858770128e-05, "loss": 3.0624, "step": 1215 }, { "epoch": 0.01, "grad_norm": 1.5529893636703491, "learning_rate": 9.998068672607146e-05, "loss": 2.8786, "step": 1220 }, { "epoch": 0.01, "grad_norm": 2.8518271446228027, "learning_rate": 9.998051391244155e-05, "loss": 2.9093, "step": 1225 }, { "epoch": 0.01, "grad_norm": 2.0863451957702637, "learning_rate": 9.998034032924092e-05, "loss": 2.8575, "step": 1230 }, { "epoch": 0.01, "grad_norm": 2.118622064590454, "learning_rate": 9.998016597647225e-05, "loss": 2.9564, "step": 1235 }, { "epoch": 0.01, "grad_norm": 1.765393853187561, "learning_rate": 9.997999085413823e-05, "loss": 2.8224, "step": 1240 }, { "epoch": 0.01, "grad_norm": 1.5686383247375488, "learning_rate": 9.997981496224153e-05, "loss": 2.749, "step": 1245 }, { "epoch": 0.01, "grad_norm": 1.6873877048492432, "learning_rate": 9.99796383007849e-05, "loss": 2.8438, "step": 1250 }, { "epoch": 0.01, "grad_norm": 3.3105854988098145, "learning_rate": 9.997946086977103e-05, "loss": 2.8848, "step": 1255 }, { "epoch": 0.01, "grad_norm": 1.2947429418563843, "learning_rate": 9.997928266920266e-05, "loss": 2.7872, "step": 1260 }, { "epoch": 0.01, "grad_norm": 1.2718141078948975, "learning_rate": 9.997910369908254e-05, "loss": 2.769, "step": 1265 }, { "epoch": 0.01, "grad_norm": 3.0559957027435303, "learning_rate": 9.99789239594134e-05, "loss": 2.8867, "step": 1270 }, { "epoch": 0.01, "grad_norm": 1.4887027740478516, "learning_rate": 9.997874345019805e-05, "loss": 2.8625, "step": 1275 }, { "epoch": 0.01, "grad_norm": 115.72406005859375, "learning_rate": 9.997856217143921e-05, "loss": 2.9906, "step": 1280 }, { "epoch": 0.01, "grad_norm": 0.9890400171279907, "learning_rate": 9.997838012313974e-05, "loss": 2.8101, "step": 1285 }, { "epoch": 0.01, "grad_norm": 1.6822614669799805, "learning_rate": 9.99781973053024e-05, "loss": 2.8012, "step": 1290 }, { "epoch": 0.01, "grad_norm": 2.7883646488189697, "learning_rate": 9.997801371793001e-05, "loss": 2.7352, "step": 1295 }, { "epoch": 0.01, "grad_norm": 1.396718144416809, "learning_rate": 9.997782936102542e-05, "loss": 2.8148, "step": 1300 }, { "epoch": 0.01, "grad_norm": 1.4444081783294678, "learning_rate": 9.997764423459144e-05, "loss": 2.8593, "step": 1305 }, { "epoch": 0.01, "grad_norm": 1.4726078510284424, "learning_rate": 9.997745833863092e-05, "loss": 2.8469, "step": 1310 }, { "epoch": 0.01, "grad_norm": 1.3248264789581299, "learning_rate": 9.997727167314676e-05, "loss": 2.7764, "step": 1315 }, { "epoch": 0.01, "grad_norm": 1.981975793838501, "learning_rate": 9.997708423814178e-05, "loss": 2.8581, "step": 1320 }, { "epoch": 0.01, "grad_norm": 1.2294409275054932, "learning_rate": 9.99768960336189e-05, "loss": 2.7558, "step": 1325 }, { "epoch": 0.01, "grad_norm": 3.829655170440674, "learning_rate": 9.997670705958102e-05, "loss": 2.7327, "step": 1330 }, { "epoch": 0.01, "grad_norm": 1.041372299194336, "learning_rate": 9.997651731603104e-05, "loss": 2.7715, "step": 1335 }, { "epoch": 0.01, "grad_norm": 1.028016448020935, "learning_rate": 9.997632680297188e-05, "loss": 2.7432, "step": 1340 }, { "epoch": 0.01, "grad_norm": 0.9429539442062378, "learning_rate": 9.997613552040648e-05, "loss": 2.8278, "step": 1345 }, { "epoch": 0.01, "grad_norm": 1.7190970182418823, "learning_rate": 9.997594346833779e-05, "loss": 2.817, "step": 1350 }, { "epoch": 0.01, "grad_norm": 1.5575354099273682, "learning_rate": 9.997575064676875e-05, "loss": 2.9085, "step": 1355 }, { "epoch": 0.01, "grad_norm": 1.1152437925338745, "learning_rate": 9.997555705570232e-05, "loss": 2.8598, "step": 1360 }, { "epoch": 0.01, "grad_norm": 1.1128276586532593, "learning_rate": 9.997536269514152e-05, "loss": 2.8496, "step": 1365 }, { "epoch": 0.01, "grad_norm": 1.4329544305801392, "learning_rate": 9.99751675650893e-05, "loss": 2.8285, "step": 1370 }, { "epoch": 0.01, "grad_norm": 2.1554365158081055, "learning_rate": 9.99749716655487e-05, "loss": 2.8921, "step": 1375 }, { "epoch": 0.01, "grad_norm": 1.9278274774551392, "learning_rate": 9.997477499652273e-05, "loss": 2.7791, "step": 1380 }, { "epoch": 0.01, "grad_norm": 15.518762588500977, "learning_rate": 9.99745775580144e-05, "loss": 2.9343, "step": 1385 }, { "epoch": 0.01, "grad_norm": 1.3025250434875488, "learning_rate": 9.997437935002676e-05, "loss": 2.8323, "step": 1390 }, { "epoch": 0.01, "grad_norm": 50.820838928222656, "learning_rate": 9.997418037256285e-05, "loss": 2.8864, "step": 1395 }, { "epoch": 0.01, "grad_norm": 1.0735454559326172, "learning_rate": 9.997398062562575e-05, "loss": 2.7584, "step": 1400 }, { "epoch": 0.01, "grad_norm": 1.7207151651382446, "learning_rate": 9.997378010921856e-05, "loss": 2.8383, "step": 1405 }, { "epoch": 0.01, "grad_norm": 1.4848235845565796, "learning_rate": 9.99735788233443e-05, "loss": 2.7149, "step": 1410 }, { "epoch": 0.01, "grad_norm": 8.94241714477539, "learning_rate": 9.997337676800613e-05, "loss": 2.831, "step": 1415 }, { "epoch": 0.01, "grad_norm": 24.95887565612793, "learning_rate": 9.997317394320713e-05, "loss": 2.7583, "step": 1420 }, { "epoch": 0.01, "grad_norm": 1.717126727104187, "learning_rate": 9.997297034895044e-05, "loss": 2.8597, "step": 1425 }, { "epoch": 0.01, "grad_norm": 4.229452610015869, "learning_rate": 9.99727659852392e-05, "loss": 2.7903, "step": 1430 }, { "epoch": 0.01, "grad_norm": 2.428849458694458, "learning_rate": 9.997256085207652e-05, "loss": 2.7342, "step": 1435 }, { "epoch": 0.01, "grad_norm": 1.4848405122756958, "learning_rate": 9.997235494946557e-05, "loss": 2.7706, "step": 1440 }, { "epoch": 0.01, "grad_norm": 1.7203773260116577, "learning_rate": 9.997214827740955e-05, "loss": 2.7191, "step": 1445 }, { "epoch": 0.01, "grad_norm": 1.9720540046691895, "learning_rate": 9.997194083591164e-05, "loss": 2.7447, "step": 1450 }, { "epoch": 0.01, "grad_norm": 1.6448341608047485, "learning_rate": 9.9971732624975e-05, "loss": 2.8033, "step": 1455 }, { "epoch": 0.01, "grad_norm": 2.6467297077178955, "learning_rate": 9.997152364460287e-05, "loss": 2.7644, "step": 1460 }, { "epoch": 0.01, "grad_norm": 1.2054628133773804, "learning_rate": 9.997131389479844e-05, "loss": 2.7555, "step": 1465 }, { "epoch": 0.01, "grad_norm": 1.8165096044540405, "learning_rate": 9.997110337556494e-05, "loss": 2.7757, "step": 1470 }, { "epoch": 0.01, "grad_norm": 1.1317425966262817, "learning_rate": 9.997089208690564e-05, "loss": 2.7829, "step": 1475 }, { "epoch": 0.01, "grad_norm": 0.990829348564148, "learning_rate": 9.997068002882377e-05, "loss": 2.7254, "step": 1480 }, { "epoch": 0.01, "grad_norm": 2.8390471935272217, "learning_rate": 9.997046720132261e-05, "loss": 2.7879, "step": 1485 }, { "epoch": 0.01, "grad_norm": 23.07222557067871, "learning_rate": 9.997025360440541e-05, "loss": 2.7923, "step": 1490 }, { "epoch": 0.01, "grad_norm": 0.7211391925811768, "learning_rate": 9.997003923807551e-05, "loss": 2.6968, "step": 1495 }, { "epoch": 0.01, "grad_norm": 1.4398927688598633, "learning_rate": 9.996982410233615e-05, "loss": 2.7873, "step": 1500 }, { "epoch": 0.01, "grad_norm": 50.704010009765625, "learning_rate": 9.996960819719068e-05, "loss": 2.7127, "step": 1505 }, { "epoch": 0.01, "grad_norm": 2.206425428390503, "learning_rate": 9.996939152264242e-05, "loss": 2.8082, "step": 1510 }, { "epoch": 0.01, "grad_norm": 28.88568115234375, "learning_rate": 9.99691740786947e-05, "loss": 2.8317, "step": 1515 }, { "epoch": 0.01, "grad_norm": 1.397343397140503, "learning_rate": 9.996895586535086e-05, "loss": 2.7866, "step": 1520 }, { "epoch": 0.01, "grad_norm": 7.045936107635498, "learning_rate": 9.996873688261428e-05, "loss": 2.8328, "step": 1525 }, { "epoch": 0.01, "grad_norm": 1.0708706378936768, "learning_rate": 9.996851713048831e-05, "loss": 2.7536, "step": 1530 }, { "epoch": 0.01, "grad_norm": 1.0406575202941895, "learning_rate": 9.996829660897637e-05, "loss": 2.8022, "step": 1535 }, { "epoch": 0.01, "grad_norm": 2.094155788421631, "learning_rate": 9.996807531808181e-05, "loss": 2.8335, "step": 1540 }, { "epoch": 0.01, "grad_norm": 1.4337419271469116, "learning_rate": 9.996785325780806e-05, "loss": 2.7575, "step": 1545 }, { "epoch": 0.01, "grad_norm": 1.5447794198989868, "learning_rate": 9.996763042815853e-05, "loss": 2.7506, "step": 1550 }, { "epoch": 0.01, "grad_norm": 1.6979362964630127, "learning_rate": 9.996740682913668e-05, "loss": 2.6666, "step": 1555 }, { "epoch": 0.01, "grad_norm": 2.205787181854248, "learning_rate": 9.996718246074591e-05, "loss": 2.8065, "step": 1560 }, { "epoch": 0.01, "grad_norm": 1.3902684450149536, "learning_rate": 9.99669573229897e-05, "loss": 2.6957, "step": 1565 }, { "epoch": 0.01, "grad_norm": 4.88633394241333, "learning_rate": 9.996673141587151e-05, "loss": 2.816, "step": 1570 }, { "epoch": 0.01, "grad_norm": 1.0045934915542603, "learning_rate": 9.996650473939482e-05, "loss": 2.7293, "step": 1575 }, { "epoch": 0.01, "grad_norm": 0.8540700078010559, "learning_rate": 9.996627729356314e-05, "loss": 2.7098, "step": 1580 }, { "epoch": 0.01, "grad_norm": 1.0516071319580078, "learning_rate": 9.996604907837993e-05, "loss": 2.8006, "step": 1585 }, { "epoch": 0.01, "grad_norm": 16.636394500732422, "learning_rate": 9.996582009384874e-05, "loss": 2.7974, "step": 1590 }, { "epoch": 0.01, "grad_norm": 1.7581427097320557, "learning_rate": 9.996559033997307e-05, "loss": 2.7972, "step": 1595 }, { "epoch": 0.01, "grad_norm": 1.1325987577438354, "learning_rate": 9.996535981675648e-05, "loss": 2.7938, "step": 1600 }, { "epoch": 0.01, "grad_norm": 0.8904371857643127, "learning_rate": 9.996512852420251e-05, "loss": 2.7344, "step": 1605 }, { "epoch": 0.01, "grad_norm": 7.01926326751709, "learning_rate": 9.996489646231471e-05, "loss": 2.7556, "step": 1610 }, { "epoch": 0.01, "grad_norm": 1.7268515825271606, "learning_rate": 9.996466363109667e-05, "loss": 2.7506, "step": 1615 }, { "epoch": 0.01, "grad_norm": 0.9160427451133728, "learning_rate": 9.996443003055196e-05, "loss": 2.7401, "step": 1620 }, { "epoch": 0.01, "grad_norm": 19.764371871948242, "learning_rate": 9.99641956606842e-05, "loss": 2.7337, "step": 1625 }, { "epoch": 0.01, "grad_norm": 1.1292012929916382, "learning_rate": 9.996396052149698e-05, "loss": 2.8114, "step": 1630 }, { "epoch": 0.01, "grad_norm": 26.08550453186035, "learning_rate": 9.996372461299392e-05, "loss": 2.7463, "step": 1635 }, { "epoch": 0.01, "grad_norm": 3.6333162784576416, "learning_rate": 9.996348793517866e-05, "loss": 2.7631, "step": 1640 }, { "epoch": 0.01, "grad_norm": 5.4758429527282715, "learning_rate": 9.996325048805484e-05, "loss": 2.6822, "step": 1645 }, { "epoch": 0.01, "grad_norm": 1.1958929300308228, "learning_rate": 9.996301227162612e-05, "loss": 2.7194, "step": 1650 }, { "epoch": 0.01, "grad_norm": 1.6850923299789429, "learning_rate": 9.996277328589618e-05, "loss": 2.7674, "step": 1655 }, { "epoch": 0.01, "grad_norm": 1.5674161911010742, "learning_rate": 9.996253353086866e-05, "loss": 2.6645, "step": 1660 }, { "epoch": 0.01, "grad_norm": 1.015993595123291, "learning_rate": 9.996229300654729e-05, "loss": 2.7631, "step": 1665 }, { "epoch": 0.01, "grad_norm": 1.671878695487976, "learning_rate": 9.996205171293576e-05, "loss": 2.7063, "step": 1670 }, { "epoch": 0.01, "grad_norm": 0.9073765277862549, "learning_rate": 9.996180965003778e-05, "loss": 2.748, "step": 1675 }, { "epoch": 0.01, "grad_norm": 1.0207575559616089, "learning_rate": 9.996156681785708e-05, "loss": 2.6792, "step": 1680 }, { "epoch": 0.01, "grad_norm": 6.360818862915039, "learning_rate": 9.996132321639741e-05, "loss": 2.7204, "step": 1685 }, { "epoch": 0.01, "grad_norm": 11.772356033325195, "learning_rate": 9.996107884566252e-05, "loss": 2.7291, "step": 1690 }, { "epoch": 0.01, "grad_norm": 0.7884344458580017, "learning_rate": 9.996083370565616e-05, "loss": 2.7181, "step": 1695 }, { "epoch": 0.01, "grad_norm": 3.1400108337402344, "learning_rate": 9.99605877963821e-05, "loss": 2.8154, "step": 1700 }, { "epoch": 0.01, "grad_norm": 0.8210126757621765, "learning_rate": 9.996034111784413e-05, "loss": 2.5414, "step": 1705 }, { "epoch": 0.01, "grad_norm": 2.3623664379119873, "learning_rate": 9.996009367004607e-05, "loss": 2.6604, "step": 1710 }, { "epoch": 0.01, "grad_norm": 6.6065144538879395, "learning_rate": 9.995984545299171e-05, "loss": 2.8238, "step": 1715 }, { "epoch": 0.01, "grad_norm": 0.817226767539978, "learning_rate": 9.995959646668488e-05, "loss": 2.806, "step": 1720 }, { "epoch": 0.01, "grad_norm": 1.2051303386688232, "learning_rate": 9.995934671112941e-05, "loss": 2.703, "step": 1725 }, { "epoch": 0.01, "grad_norm": 1.3889830112457275, "learning_rate": 9.995909618632915e-05, "loss": 2.7129, "step": 1730 }, { "epoch": 0.01, "grad_norm": 1.2944672107696533, "learning_rate": 9.995884489228794e-05, "loss": 2.7116, "step": 1735 }, { "epoch": 0.01, "grad_norm": 3.228343963623047, "learning_rate": 9.995859282900967e-05, "loss": 2.7542, "step": 1740 }, { "epoch": 0.01, "grad_norm": 1.6474348306655884, "learning_rate": 9.995833999649821e-05, "loss": 2.7742, "step": 1745 }, { "epoch": 0.01, "grad_norm": 1.3084781169891357, "learning_rate": 9.995808639475746e-05, "loss": 2.8306, "step": 1750 }, { "epoch": 0.01, "grad_norm": 1.007794976234436, "learning_rate": 9.995783202379131e-05, "loss": 2.7374, "step": 1755 }, { "epoch": 0.01, "grad_norm": 0.917638897895813, "learning_rate": 9.99575768836037e-05, "loss": 2.7212, "step": 1760 }, { "epoch": 0.01, "grad_norm": 1.4690898656845093, "learning_rate": 9.995732097419855e-05, "loss": 2.7358, "step": 1765 }, { "epoch": 0.01, "grad_norm": 7.670183181762695, "learning_rate": 9.995706429557978e-05, "loss": 2.6997, "step": 1770 }, { "epoch": 0.01, "grad_norm": 1.3014492988586426, "learning_rate": 9.995680684775137e-05, "loss": 2.7378, "step": 1775 }, { "epoch": 0.01, "grad_norm": 5.758452892303467, "learning_rate": 9.995654863071729e-05, "loss": 2.7555, "step": 1780 }, { "epoch": 0.01, "grad_norm": 2.1906275749206543, "learning_rate": 9.995628964448147e-05, "loss": 2.5928, "step": 1785 }, { "epoch": 0.01, "grad_norm": 1.106614589691162, "learning_rate": 9.995602988904793e-05, "loss": 2.802, "step": 1790 }, { "epoch": 0.01, "grad_norm": 5.470571517944336, "learning_rate": 9.995576936442068e-05, "loss": 2.7257, "step": 1795 }, { "epoch": 0.01, "grad_norm": 2.212959051132202, "learning_rate": 9.995550807060371e-05, "loss": 2.7024, "step": 1800 }, { "epoch": 0.01, "grad_norm": 0.7684453725814819, "learning_rate": 9.995524600760105e-05, "loss": 2.7923, "step": 1805 }, { "epoch": 0.01, "grad_norm": 0.8690535426139832, "learning_rate": 9.995498317541674e-05, "loss": 2.6878, "step": 1810 }, { "epoch": 0.01, "grad_norm": 4.859379768371582, "learning_rate": 9.995471957405483e-05, "loss": 2.7151, "step": 1815 }, { "epoch": 0.01, "grad_norm": 1.1911873817443848, "learning_rate": 9.995445520351935e-05, "loss": 2.7114, "step": 1820 }, { "epoch": 0.01, "grad_norm": 1.571197748184204, "learning_rate": 9.995419006381442e-05, "loss": 2.6384, "step": 1825 }, { "epoch": 0.01, "grad_norm": 9.359936714172363, "learning_rate": 9.995392415494407e-05, "loss": 2.7371, "step": 1830 }, { "epoch": 0.01, "grad_norm": 2.551485776901245, "learning_rate": 9.995365747691243e-05, "loss": 2.8073, "step": 1835 }, { "epoch": 0.01, "grad_norm": 1.3948525190353394, "learning_rate": 9.995339002972359e-05, "loss": 2.6985, "step": 1840 }, { "epoch": 0.01, "grad_norm": 1.2203726768493652, "learning_rate": 9.995312181338168e-05, "loss": 2.7357, "step": 1845 }, { "epoch": 0.01, "grad_norm": 0.8107905387878418, "learning_rate": 9.995285282789082e-05, "loss": 2.659, "step": 1850 }, { "epoch": 0.01, "grad_norm": 0.9072896838188171, "learning_rate": 9.995258307325515e-05, "loss": 2.6539, "step": 1855 }, { "epoch": 0.01, "grad_norm": 2.276167154312134, "learning_rate": 9.995231254947883e-05, "loss": 2.6353, "step": 1860 }, { "epoch": 0.01, "grad_norm": 7.797708511352539, "learning_rate": 9.995204125656603e-05, "loss": 2.7055, "step": 1865 }, { "epoch": 0.01, "grad_norm": 0.6722606420516968, "learning_rate": 9.995176919452092e-05, "loss": 2.7189, "step": 1870 }, { "epoch": 0.01, "grad_norm": 0.9699838161468506, "learning_rate": 9.995149636334769e-05, "loss": 2.7715, "step": 1875 }, { "epoch": 0.01, "grad_norm": 1.102444052696228, "learning_rate": 9.995122276305054e-05, "loss": 2.6498, "step": 1880 }, { "epoch": 0.01, "grad_norm": 2.178607940673828, "learning_rate": 9.995094839363368e-05, "loss": 2.6607, "step": 1885 }, { "epoch": 0.01, "grad_norm": 1.7641550302505493, "learning_rate": 9.995067325510132e-05, "loss": 2.6827, "step": 1890 }, { "epoch": 0.01, "grad_norm": 1.5672744512557983, "learning_rate": 9.995039734745774e-05, "loss": 2.7219, "step": 1895 }, { "epoch": 0.01, "grad_norm": 8.507780075073242, "learning_rate": 9.995012067070715e-05, "loss": 2.7212, "step": 1900 }, { "epoch": 0.02, "grad_norm": 1.417026162147522, "learning_rate": 9.994984322485383e-05, "loss": 2.6795, "step": 1905 }, { "epoch": 0.02, "grad_norm": 0.9245043396949768, "learning_rate": 9.994956500990205e-05, "loss": 2.7639, "step": 1910 }, { "epoch": 0.02, "grad_norm": 1.034032940864563, "learning_rate": 9.994928602585608e-05, "loss": 2.6581, "step": 1915 }, { "epoch": 0.02, "grad_norm": 1.5898463726043701, "learning_rate": 9.994900627272023e-05, "loss": 2.7165, "step": 1920 }, { "epoch": 0.02, "grad_norm": 1.3166590929031372, "learning_rate": 9.99487257504988e-05, "loss": 2.7221, "step": 1925 }, { "epoch": 0.02, "grad_norm": 2.544686794281006, "learning_rate": 9.994844445919612e-05, "loss": 2.695, "step": 1930 }, { "epoch": 0.02, "grad_norm": 0.8995977640151978, "learning_rate": 9.994816239881648e-05, "loss": 2.655, "step": 1935 }, { "epoch": 0.02, "grad_norm": 7.094485759735107, "learning_rate": 9.99478795693643e-05, "loss": 2.5798, "step": 1940 }, { "epoch": 0.02, "grad_norm": 1.2805448770523071, "learning_rate": 9.994759597084384e-05, "loss": 2.6857, "step": 1945 }, { "epoch": 0.02, "grad_norm": 1.4689363241195679, "learning_rate": 9.994731160325955e-05, "loss": 2.7157, "step": 1950 }, { "epoch": 0.02, "grad_norm": 0.7996886968612671, "learning_rate": 9.994702646661577e-05, "loss": 2.6362, "step": 1955 }, { "epoch": 0.02, "grad_norm": 0.8204803466796875, "learning_rate": 9.994674056091689e-05, "loss": 2.6086, "step": 1960 }, { "epoch": 0.02, "grad_norm": 1.5984420776367188, "learning_rate": 9.994645388616732e-05, "loss": 2.6015, "step": 1965 }, { "epoch": 0.02, "grad_norm": 1.5654000043869019, "learning_rate": 9.994616644237148e-05, "loss": 2.7821, "step": 1970 }, { "epoch": 0.02, "grad_norm": 0.9124917984008789, "learning_rate": 9.994587822953378e-05, "loss": 2.6686, "step": 1975 }, { "epoch": 0.02, "grad_norm": 0.7694497108459473, "learning_rate": 9.994558924765866e-05, "loss": 2.6237, "step": 1980 }, { "epoch": 0.02, "grad_norm": 1.5341767072677612, "learning_rate": 9.99452994967506e-05, "loss": 2.7778, "step": 1985 }, { "epoch": 0.02, "grad_norm": 1.4850653409957886, "learning_rate": 9.994500897681401e-05, "loss": 2.7589, "step": 1990 }, { "epoch": 0.02, "grad_norm": 2.4870846271514893, "learning_rate": 9.994471768785339e-05, "loss": 2.7315, "step": 1995 }, { "epoch": 0.02, "grad_norm": 1.4255197048187256, "learning_rate": 9.994442562987324e-05, "loss": 2.7013, "step": 2000 }, { "epoch": 0.02, "grad_norm": 1.0722061395645142, "learning_rate": 9.994413280287803e-05, "loss": 2.7133, "step": 2005 }, { "epoch": 0.02, "grad_norm": 1.4606610536575317, "learning_rate": 9.994383920687229e-05, "loss": 2.683, "step": 2010 }, { "epoch": 0.02, "grad_norm": 0.9361623525619507, "learning_rate": 9.994354484186053e-05, "loss": 2.7437, "step": 2015 }, { "epoch": 0.02, "grad_norm": 22.254789352416992, "learning_rate": 9.994324970784727e-05, "loss": 2.7167, "step": 2020 }, { "epoch": 0.02, "grad_norm": 1.071345329284668, "learning_rate": 9.994295380483709e-05, "loss": 2.7143, "step": 2025 }, { "epoch": 0.02, "grad_norm": 0.9171513319015503, "learning_rate": 9.99426571328345e-05, "loss": 2.7192, "step": 2030 }, { "epoch": 0.02, "grad_norm": 9.528144836425781, "learning_rate": 9.994235969184411e-05, "loss": 2.6516, "step": 2035 }, { "epoch": 0.02, "grad_norm": 1.8615336418151855, "learning_rate": 9.994206148187046e-05, "loss": 2.6798, "step": 2040 }, { "epoch": 0.02, "grad_norm": 34.853153228759766, "learning_rate": 9.994176250291818e-05, "loss": 2.7852, "step": 2045 }, { "epoch": 0.02, "grad_norm": 0.697297215461731, "learning_rate": 9.994146275499186e-05, "loss": 2.752, "step": 2050 }, { "epoch": 0.02, "grad_norm": 1.4314467906951904, "learning_rate": 9.994116223809611e-05, "loss": 2.5762, "step": 2055 }, { "epoch": 0.02, "grad_norm": 1.0536082983016968, "learning_rate": 9.994086095223557e-05, "loss": 2.6874, "step": 2060 }, { "epoch": 0.02, "grad_norm": 6.412439823150635, "learning_rate": 9.994055889741485e-05, "loss": 2.6833, "step": 2065 }, { "epoch": 0.02, "grad_norm": 1.0813783407211304, "learning_rate": 9.994025607363862e-05, "loss": 2.6883, "step": 2070 }, { "epoch": 0.02, "grad_norm": 0.8327749371528625, "learning_rate": 9.993995248091155e-05, "loss": 2.5716, "step": 2075 }, { "epoch": 0.02, "grad_norm": 0.7512326836585999, "learning_rate": 9.993964811923831e-05, "loss": 2.6846, "step": 2080 }, { "epoch": 0.02, "grad_norm": 6.415153503417969, "learning_rate": 9.993934298862358e-05, "loss": 2.6647, "step": 2085 }, { "epoch": 0.02, "grad_norm": 0.9639997482299805, "learning_rate": 9.993903708907205e-05, "loss": 2.6004, "step": 2090 }, { "epoch": 0.02, "grad_norm": 1.4142882823944092, "learning_rate": 9.993873042058845e-05, "loss": 2.5658, "step": 2095 }, { "epoch": 0.02, "grad_norm": 1.1546950340270996, "learning_rate": 9.993842298317749e-05, "loss": 2.6366, "step": 2100 }, { "epoch": 0.02, "grad_norm": 1.0131210088729858, "learning_rate": 9.993811477684392e-05, "loss": 2.7111, "step": 2105 }, { "epoch": 0.02, "grad_norm": 1.1663283109664917, "learning_rate": 9.993780580159246e-05, "loss": 2.5807, "step": 2110 }, { "epoch": 0.02, "grad_norm": 0.9920238852500916, "learning_rate": 9.993749605742788e-05, "loss": 2.6535, "step": 2115 }, { "epoch": 0.02, "grad_norm": 1.2432554960250854, "learning_rate": 9.993718554435495e-05, "loss": 2.5842, "step": 2120 }, { "epoch": 0.02, "grad_norm": 12.608223915100098, "learning_rate": 9.993687426237846e-05, "loss": 2.5905, "step": 2125 }, { "epoch": 0.02, "grad_norm": 2.0823986530303955, "learning_rate": 9.993656221150318e-05, "loss": 2.6202, "step": 2130 }, { "epoch": 0.02, "grad_norm": 1.9012675285339355, "learning_rate": 9.993624939173394e-05, "loss": 2.6793, "step": 2135 }, { "epoch": 0.02, "grad_norm": 2.0575335025787354, "learning_rate": 9.993593580307553e-05, "loss": 2.6589, "step": 2140 }, { "epoch": 0.02, "grad_norm": 4.556035995483398, "learning_rate": 9.99356214455328e-05, "loss": 2.7095, "step": 2145 }, { "epoch": 0.02, "grad_norm": 0.8742093443870544, "learning_rate": 9.993530631911059e-05, "loss": 2.6921, "step": 2150 }, { "epoch": 0.02, "grad_norm": 2.047661066055298, "learning_rate": 9.993499042381373e-05, "loss": 2.714, "step": 2155 }, { "epoch": 0.02, "grad_norm": 5.008732318878174, "learning_rate": 9.993467375964711e-05, "loss": 2.6199, "step": 2160 }, { "epoch": 0.02, "grad_norm": 0.8367443084716797, "learning_rate": 9.99343563266156e-05, "loss": 2.7031, "step": 2165 }, { "epoch": 0.02, "grad_norm": 1.1222851276397705, "learning_rate": 9.993403812472408e-05, "loss": 2.696, "step": 2170 }, { "epoch": 0.02, "grad_norm": 2.223984956741333, "learning_rate": 9.993371915397745e-05, "loss": 2.6867, "step": 2175 }, { "epoch": 0.02, "grad_norm": 0.9649681448936462, "learning_rate": 9.993339941438062e-05, "loss": 2.7015, "step": 2180 }, { "epoch": 0.02, "grad_norm": 1.5247883796691895, "learning_rate": 9.993307890593853e-05, "loss": 2.6557, "step": 2185 }, { "epoch": 0.02, "grad_norm": 0.8864256143569946, "learning_rate": 9.993275762865609e-05, "loss": 2.5676, "step": 2190 }, { "epoch": 0.02, "grad_norm": 1.3801544904708862, "learning_rate": 9.993243558253826e-05, "loss": 2.6614, "step": 2195 }, { "epoch": 0.02, "grad_norm": 0.7034408450126648, "learning_rate": 9.993211276759e-05, "loss": 2.7417, "step": 2200 }, { "epoch": 0.02, "grad_norm": 0.8581990599632263, "learning_rate": 9.993178918381628e-05, "loss": 2.596, "step": 2205 }, { "epoch": 0.02, "grad_norm": 26.700443267822266, "learning_rate": 9.993146483122207e-05, "loss": 2.6637, "step": 2210 }, { "epoch": 0.02, "grad_norm": 1.304459810256958, "learning_rate": 9.993113970981238e-05, "loss": 2.6531, "step": 2215 }, { "epoch": 0.02, "grad_norm": 1.3331429958343506, "learning_rate": 9.993081381959221e-05, "loss": 2.58, "step": 2220 }, { "epoch": 0.02, "grad_norm": 12.075164794921875, "learning_rate": 9.99304871605666e-05, "loss": 2.6854, "step": 2225 }, { "epoch": 0.02, "grad_norm": 0.7252376079559326, "learning_rate": 9.993015973274053e-05, "loss": 2.7194, "step": 2230 }, { "epoch": 0.02, "grad_norm": 1.0706911087036133, "learning_rate": 9.992983153611908e-05, "loss": 2.6831, "step": 2235 }, { "epoch": 0.02, "grad_norm": 1.0827287435531616, "learning_rate": 9.99295025707073e-05, "loss": 2.6398, "step": 2240 }, { "epoch": 0.02, "grad_norm": 3.6655635833740234, "learning_rate": 9.992917283651024e-05, "loss": 2.6816, "step": 2245 }, { "epoch": 0.02, "grad_norm": 1.4088163375854492, "learning_rate": 9.992884233353298e-05, "loss": 2.6983, "step": 2250 }, { "epoch": 0.02, "grad_norm": 0.8236873149871826, "learning_rate": 9.992851106178063e-05, "loss": 2.6672, "step": 2255 }, { "epoch": 0.02, "grad_norm": 0.7171049118041992, "learning_rate": 9.992817902125826e-05, "loss": 2.676, "step": 2260 }, { "epoch": 0.02, "grad_norm": 32.81167984008789, "learning_rate": 9.9927846211971e-05, "loss": 2.6479, "step": 2265 }, { "epoch": 0.02, "grad_norm": 0.7687113881111145, "learning_rate": 9.992751263392397e-05, "loss": 2.6112, "step": 2270 }, { "epoch": 0.02, "grad_norm": 4.405129909515381, "learning_rate": 9.992717828712232e-05, "loss": 2.6938, "step": 2275 }, { "epoch": 0.02, "grad_norm": 0.7937389016151428, "learning_rate": 9.992684317157117e-05, "loss": 2.585, "step": 2280 }, { "epoch": 0.02, "grad_norm": 1.074177861213684, "learning_rate": 9.992650728727571e-05, "loss": 2.6398, "step": 2285 }, { "epoch": 0.02, "grad_norm": 8.517021179199219, "learning_rate": 9.99261706342411e-05, "loss": 2.6995, "step": 2290 }, { "epoch": 0.02, "grad_norm": 1.3561235666275024, "learning_rate": 9.99258332124725e-05, "loss": 2.7281, "step": 2295 }, { "epoch": 0.02, "grad_norm": 5.010215759277344, "learning_rate": 9.992549502197514e-05, "loss": 2.6056, "step": 2300 }, { "epoch": 0.02, "grad_norm": 7.4411301612854, "learning_rate": 9.992515606275423e-05, "loss": 2.6041, "step": 2305 }, { "epoch": 0.02, "grad_norm": 0.7985312342643738, "learning_rate": 9.992481633481495e-05, "loss": 2.6416, "step": 2310 }, { "epoch": 0.02, "grad_norm": 6.46704626083374, "learning_rate": 9.992447583816257e-05, "loss": 2.7358, "step": 2315 }, { "epoch": 0.02, "grad_norm": 0.8527347445487976, "learning_rate": 9.992413457280231e-05, "loss": 2.6514, "step": 2320 }, { "epoch": 0.02, "grad_norm": 1.2956560850143433, "learning_rate": 9.992379253873943e-05, "loss": 2.5756, "step": 2325 }, { "epoch": 0.02, "grad_norm": 0.9508882164955139, "learning_rate": 9.99234497359792e-05, "loss": 2.6841, "step": 2330 }, { "epoch": 0.02, "grad_norm": 6.355693340301514, "learning_rate": 9.992310616452689e-05, "loss": 2.5929, "step": 2335 }, { "epoch": 0.02, "grad_norm": 0.7363661527633667, "learning_rate": 9.99227618243878e-05, "loss": 2.6643, "step": 2340 }, { "epoch": 0.02, "grad_norm": 1.2793762683868408, "learning_rate": 9.992241671556724e-05, "loss": 2.6497, "step": 2345 }, { "epoch": 0.02, "grad_norm": 36.86465835571289, "learning_rate": 9.99220708380705e-05, "loss": 2.7223, "step": 2350 }, { "epoch": 0.02, "grad_norm": 1.054565191268921, "learning_rate": 9.99217241919029e-05, "loss": 2.6423, "step": 2355 }, { "epoch": 0.02, "grad_norm": 12.177027702331543, "learning_rate": 9.992137677706982e-05, "loss": 2.6634, "step": 2360 }, { "epoch": 0.02, "grad_norm": 1.2533836364746094, "learning_rate": 9.99210285935766e-05, "loss": 2.6409, "step": 2365 }, { "epoch": 0.02, "grad_norm": 1.1261160373687744, "learning_rate": 9.992067964142856e-05, "loss": 2.681, "step": 2370 }, { "epoch": 0.02, "grad_norm": 0.7593612670898438, "learning_rate": 9.992032992063109e-05, "loss": 2.6487, "step": 2375 }, { "epoch": 0.02, "grad_norm": 6.753429889678955, "learning_rate": 9.99199794311896e-05, "loss": 2.6832, "step": 2380 }, { "epoch": 0.02, "grad_norm": 1.1274075508117676, "learning_rate": 9.991962817310947e-05, "loss": 2.6736, "step": 2385 }, { "epoch": 0.02, "grad_norm": 0.8044716715812683, "learning_rate": 9.991927614639612e-05, "loss": 2.6501, "step": 2390 }, { "epoch": 0.02, "grad_norm": 12.255712509155273, "learning_rate": 9.991892335105493e-05, "loss": 2.5855, "step": 2395 }, { "epoch": 0.02, "grad_norm": 0.6356988549232483, "learning_rate": 9.991856978709137e-05, "loss": 2.679, "step": 2400 }, { "epoch": 0.02, "grad_norm": 7.324370384216309, "learning_rate": 9.99182154545109e-05, "loss": 2.6523, "step": 2405 }, { "epoch": 0.02, "grad_norm": 1.7722768783569336, "learning_rate": 9.991786035331893e-05, "loss": 2.6517, "step": 2410 }, { "epoch": 0.02, "grad_norm": 1.3676203489303589, "learning_rate": 9.991750448352096e-05, "loss": 2.5868, "step": 2415 }, { "epoch": 0.02, "grad_norm": 1.0643068552017212, "learning_rate": 9.991714784512243e-05, "loss": 2.5688, "step": 2420 }, { "epoch": 0.02, "grad_norm": 0.7412228584289551, "learning_rate": 9.991679043812887e-05, "loss": 2.6014, "step": 2425 }, { "epoch": 0.02, "grad_norm": 5.163914203643799, "learning_rate": 9.99164322625458e-05, "loss": 2.6361, "step": 2430 }, { "epoch": 0.02, "grad_norm": 1.0220412015914917, "learning_rate": 9.991607331837867e-05, "loss": 2.6283, "step": 2435 }, { "epoch": 0.02, "grad_norm": 27.322858810424805, "learning_rate": 9.991571360563306e-05, "loss": 2.673, "step": 2440 }, { "epoch": 0.02, "grad_norm": 2.466646194458008, "learning_rate": 9.99153531243145e-05, "loss": 2.6234, "step": 2445 }, { "epoch": 0.02, "grad_norm": 4.844601154327393, "learning_rate": 9.991499187442852e-05, "loss": 2.5614, "step": 2450 }, { "epoch": 0.02, "grad_norm": 1.1777429580688477, "learning_rate": 9.99146298559807e-05, "loss": 2.5802, "step": 2455 }, { "epoch": 0.02, "grad_norm": 0.8394126892089844, "learning_rate": 9.991426706897661e-05, "loss": 2.5771, "step": 2460 }, { "epoch": 0.02, "grad_norm": 4.4165449142456055, "learning_rate": 9.991390351342185e-05, "loss": 2.5206, "step": 2465 }, { "epoch": 0.02, "grad_norm": 0.7794109582901001, "learning_rate": 9.991353918932198e-05, "loss": 2.611, "step": 2470 }, { "epoch": 0.02, "grad_norm": 1.260448932647705, "learning_rate": 9.991317409668264e-05, "loss": 2.6083, "step": 2475 }, { "epoch": 0.02, "grad_norm": 3.569079637527466, "learning_rate": 9.991280823550945e-05, "loss": 2.6527, "step": 2480 }, { "epoch": 0.02, "grad_norm": 5.893279075622559, "learning_rate": 9.991244160580803e-05, "loss": 2.6463, "step": 2485 }, { "epoch": 0.02, "grad_norm": 1.4530965089797974, "learning_rate": 9.991207420758404e-05, "loss": 2.6654, "step": 2490 }, { "epoch": 0.02, "grad_norm": 1.257289171218872, "learning_rate": 9.991170604084314e-05, "loss": 2.6522, "step": 2495 }, { "epoch": 0.02, "grad_norm": 2.486673593521118, "learning_rate": 9.991133710559097e-05, "loss": 2.6499, "step": 2500 }, { "epoch": 0.02, "grad_norm": 0.8248240351676941, "learning_rate": 9.991096740183324e-05, "loss": 2.6071, "step": 2505 }, { "epoch": 0.02, "grad_norm": 1.0029692649841309, "learning_rate": 9.991059692957563e-05, "loss": 2.606, "step": 2510 }, { "epoch": 0.02, "grad_norm": 1.1006557941436768, "learning_rate": 9.991022568882384e-05, "loss": 2.5984, "step": 2515 }, { "epoch": 0.02, "grad_norm": 0.9511757493019104, "learning_rate": 9.99098536795836e-05, "loss": 2.5683, "step": 2520 }, { "epoch": 0.02, "grad_norm": 5.538845062255859, "learning_rate": 9.990948090186063e-05, "loss": 2.6501, "step": 2525 }, { "epoch": 0.02, "grad_norm": 1.8992868661880493, "learning_rate": 9.990910735566065e-05, "loss": 2.5659, "step": 2530 }, { "epoch": 0.02, "grad_norm": 1.5456171035766602, "learning_rate": 9.990873304098946e-05, "loss": 2.6341, "step": 2535 }, { "epoch": 0.02, "grad_norm": 0.8612980246543884, "learning_rate": 9.990835795785278e-05, "loss": 2.5551, "step": 2540 }, { "epoch": 0.02, "grad_norm": 1.130746841430664, "learning_rate": 9.990798210625642e-05, "loss": 2.5594, "step": 2545 }, { "epoch": 0.02, "grad_norm": 1.2934924364089966, "learning_rate": 9.990760548620613e-05, "loss": 2.6087, "step": 2550 }, { "epoch": 0.02, "grad_norm": 4.050301551818848, "learning_rate": 9.990722809770773e-05, "loss": 2.6236, "step": 2555 }, { "epoch": 0.02, "grad_norm": 0.729785680770874, "learning_rate": 9.990684994076702e-05, "loss": 2.6244, "step": 2560 }, { "epoch": 0.02, "grad_norm": 1.145810604095459, "learning_rate": 9.990647101538983e-05, "loss": 2.6555, "step": 2565 }, { "epoch": 0.02, "grad_norm": 1.6413639783859253, "learning_rate": 9.990609132158201e-05, "loss": 2.64, "step": 2570 }, { "epoch": 0.02, "grad_norm": 2.1194324493408203, "learning_rate": 9.990571085934938e-05, "loss": 2.6027, "step": 2575 }, { "epoch": 0.02, "grad_norm": 0.8772727847099304, "learning_rate": 9.990532962869783e-05, "loss": 2.6206, "step": 2580 }, { "epoch": 0.02, "grad_norm": 1.234667181968689, "learning_rate": 9.990494762963318e-05, "loss": 2.5943, "step": 2585 }, { "epoch": 0.02, "grad_norm": 1.3256193399429321, "learning_rate": 9.990456486216136e-05, "loss": 2.6257, "step": 2590 }, { "epoch": 0.02, "grad_norm": 2.212775230407715, "learning_rate": 9.990418132628824e-05, "loss": 2.6814, "step": 2595 }, { "epoch": 0.02, "grad_norm": 1.8071650266647339, "learning_rate": 9.990379702201974e-05, "loss": 2.5729, "step": 2600 }, { "epoch": 0.02, "grad_norm": 0.791434109210968, "learning_rate": 9.990341194936173e-05, "loss": 2.7065, "step": 2605 }, { "epoch": 0.02, "grad_norm": 1.07462477684021, "learning_rate": 9.990302610832021e-05, "loss": 2.6837, "step": 2610 }, { "epoch": 0.02, "grad_norm": 1.625205397605896, "learning_rate": 9.990263949890108e-05, "loss": 2.6099, "step": 2615 }, { "epoch": 0.02, "grad_norm": 1.4051823616027832, "learning_rate": 9.990225212111029e-05, "loss": 2.5947, "step": 2620 }, { "epoch": 0.02, "grad_norm": 1.0545560121536255, "learning_rate": 9.990186397495383e-05, "loss": 2.5848, "step": 2625 }, { "epoch": 0.02, "grad_norm": 1.249193549156189, "learning_rate": 9.990147506043766e-05, "loss": 2.5816, "step": 2630 }, { "epoch": 0.02, "grad_norm": 1.0229064226150513, "learning_rate": 9.990108537756775e-05, "loss": 2.6198, "step": 2635 }, { "epoch": 0.02, "grad_norm": 0.8591925501823425, "learning_rate": 9.990069492635015e-05, "loss": 2.6176, "step": 2640 }, { "epoch": 0.02, "grad_norm": 3.251058340072632, "learning_rate": 9.990030370679082e-05, "loss": 2.606, "step": 2645 }, { "epoch": 0.02, "grad_norm": 1.0182942152023315, "learning_rate": 9.989991171889581e-05, "loss": 2.5944, "step": 2650 }, { "epoch": 0.02, "grad_norm": 1.5466761589050293, "learning_rate": 9.989951896267115e-05, "loss": 2.604, "step": 2655 }, { "epoch": 0.02, "grad_norm": 8.586068153381348, "learning_rate": 9.989912543812289e-05, "loss": 2.5963, "step": 2660 }, { "epoch": 0.02, "grad_norm": 0.9820445775985718, "learning_rate": 9.989873114525709e-05, "loss": 2.5798, "step": 2665 }, { "epoch": 0.02, "grad_norm": 1.4473347663879395, "learning_rate": 9.989833608407981e-05, "loss": 2.5893, "step": 2670 }, { "epoch": 0.02, "grad_norm": 0.9268401861190796, "learning_rate": 9.989794025459715e-05, "loss": 2.6034, "step": 2675 }, { "epoch": 0.02, "grad_norm": 0.9264147281646729, "learning_rate": 9.989754365681519e-05, "loss": 2.5833, "step": 2680 }, { "epoch": 0.02, "grad_norm": 17.666805267333984, "learning_rate": 9.989714629074006e-05, "loss": 2.7195, "step": 2685 }, { "epoch": 0.02, "grad_norm": 0.9174264669418335, "learning_rate": 9.989674815637784e-05, "loss": 2.6702, "step": 2690 }, { "epoch": 0.02, "grad_norm": 0.6662594079971313, "learning_rate": 9.989634925373468e-05, "loss": 2.5173, "step": 2695 }, { "epoch": 0.02, "grad_norm": 8.5391206741333, "learning_rate": 9.989594958281673e-05, "loss": 2.5754, "step": 2700 }, { "epoch": 0.02, "grad_norm": 0.8522529006004333, "learning_rate": 9.989554914363015e-05, "loss": 2.51, "step": 2705 }, { "epoch": 0.02, "grad_norm": 1.9496815204620361, "learning_rate": 9.989514793618109e-05, "loss": 2.5089, "step": 2710 }, { "epoch": 0.02, "grad_norm": 5.440359115600586, "learning_rate": 9.989474596047572e-05, "loss": 2.563, "step": 2715 }, { "epoch": 0.02, "grad_norm": 1.8642234802246094, "learning_rate": 9.989434321652024e-05, "loss": 2.5085, "step": 2720 }, { "epoch": 0.02, "grad_norm": 0.7927899360656738, "learning_rate": 9.989393970432085e-05, "loss": 2.6348, "step": 2725 }, { "epoch": 0.02, "grad_norm": 3.1566500663757324, "learning_rate": 9.989353542388378e-05, "loss": 2.5798, "step": 2730 }, { "epoch": 0.02, "grad_norm": 0.9913167357444763, "learning_rate": 9.989313037521524e-05, "loss": 2.62, "step": 2735 }, { "epoch": 0.02, "grad_norm": 1.1157406568527222, "learning_rate": 9.989272455832145e-05, "loss": 2.6327, "step": 2740 }, { "epoch": 0.02, "grad_norm": 1.3982287645339966, "learning_rate": 9.989231797320869e-05, "loss": 2.5861, "step": 2745 }, { "epoch": 0.02, "grad_norm": 1.3292481899261475, "learning_rate": 9.989191061988319e-05, "loss": 2.5859, "step": 2750 }, { "epoch": 0.02, "grad_norm": 0.7453621029853821, "learning_rate": 9.989150249835125e-05, "loss": 2.6094, "step": 2755 }, { "epoch": 0.02, "grad_norm": 1.9909714460372925, "learning_rate": 9.989109360861913e-05, "loss": 2.6864, "step": 2760 }, { "epoch": 0.02, "grad_norm": 0.8525038957595825, "learning_rate": 9.989068395069313e-05, "loss": 2.4956, "step": 2765 }, { "epoch": 0.02, "grad_norm": 0.8183479309082031, "learning_rate": 9.989027352457958e-05, "loss": 2.5579, "step": 2770 }, { "epoch": 0.02, "grad_norm": 0.981203019618988, "learning_rate": 9.988986233028478e-05, "loss": 2.5626, "step": 2775 }, { "epoch": 0.02, "grad_norm": 8.61716079711914, "learning_rate": 9.988945036781507e-05, "loss": 2.6281, "step": 2780 }, { "epoch": 0.02, "grad_norm": 0.7685188055038452, "learning_rate": 9.988903763717679e-05, "loss": 2.6195, "step": 2785 }, { "epoch": 0.02, "grad_norm": 0.8435307145118713, "learning_rate": 9.988862413837628e-05, "loss": 2.5109, "step": 2790 }, { "epoch": 0.02, "grad_norm": 10.595979690551758, "learning_rate": 9.988820987141993e-05, "loss": 2.5709, "step": 2795 }, { "epoch": 0.02, "grad_norm": 0.975407600402832, "learning_rate": 9.988779483631409e-05, "loss": 2.5843, "step": 2800 }, { "epoch": 0.02, "grad_norm": 0.789325475692749, "learning_rate": 9.98873790330652e-05, "loss": 2.6096, "step": 2805 }, { "epoch": 0.02, "grad_norm": 0.8474391102790833, "learning_rate": 9.988696246167961e-05, "loss": 2.5648, "step": 2810 }, { "epoch": 0.02, "grad_norm": 2.710428476333618, "learning_rate": 9.988654512216378e-05, "loss": 2.5348, "step": 2815 }, { "epoch": 0.02, "grad_norm": 0.8019130229949951, "learning_rate": 9.988612701452409e-05, "loss": 2.6554, "step": 2820 }, { "epoch": 0.02, "grad_norm": 1.5471311807632446, "learning_rate": 9.988570813876701e-05, "loss": 2.6175, "step": 2825 }, { "epoch": 0.02, "grad_norm": 1.0394319295883179, "learning_rate": 9.988528849489897e-05, "loss": 2.5762, "step": 2830 }, { "epoch": 0.02, "grad_norm": 4.228000164031982, "learning_rate": 9.988486808292646e-05, "loss": 2.5817, "step": 2835 }, { "epoch": 0.02, "grad_norm": 1.4531702995300293, "learning_rate": 9.988444690285592e-05, "loss": 2.528, "step": 2840 }, { "epoch": 0.02, "grad_norm": 0.9690269827842712, "learning_rate": 9.988402495469385e-05, "loss": 2.6308, "step": 2845 }, { "epoch": 0.02, "grad_norm": 56.95972442626953, "learning_rate": 9.988360223844673e-05, "loss": 2.5399, "step": 2850 }, { "epoch": 0.02, "grad_norm": 2.9274842739105225, "learning_rate": 9.98831787541211e-05, "loss": 2.5353, "step": 2855 }, { "epoch": 0.02, "grad_norm": 9.042009353637695, "learning_rate": 9.988275450172346e-05, "loss": 2.4595, "step": 2860 }, { "epoch": 0.02, "grad_norm": 0.9561848640441895, "learning_rate": 9.988232948126037e-05, "loss": 2.5193, "step": 2865 }, { "epoch": 0.02, "grad_norm": 0.9261096119880676, "learning_rate": 9.988190369273833e-05, "loss": 2.5824, "step": 2870 }, { "epoch": 0.02, "grad_norm": 1.8711585998535156, "learning_rate": 9.988147713616391e-05, "loss": 2.5151, "step": 2875 }, { "epoch": 0.02, "grad_norm": 0.7945968508720398, "learning_rate": 9.988104981154371e-05, "loss": 2.6443, "step": 2880 }, { "epoch": 0.02, "grad_norm": 1.706897497177124, "learning_rate": 9.988062171888429e-05, "loss": 2.4404, "step": 2885 }, { "epoch": 0.02, "grad_norm": 2.204029083251953, "learning_rate": 9.988019285819221e-05, "loss": 2.6046, "step": 2890 }, { "epoch": 0.02, "grad_norm": 2.2460718154907227, "learning_rate": 9.987976322947411e-05, "loss": 2.6448, "step": 2895 }, { "epoch": 0.02, "grad_norm": 7.980650901794434, "learning_rate": 9.98793328327366e-05, "loss": 2.6573, "step": 2900 }, { "epoch": 0.02, "grad_norm": 0.595098078250885, "learning_rate": 9.98789016679863e-05, "loss": 2.6027, "step": 2905 }, { "epoch": 0.02, "grad_norm": 0.8913105130195618, "learning_rate": 9.987846973522985e-05, "loss": 2.592, "step": 2910 }, { "epoch": 0.02, "grad_norm": 1.37417733669281, "learning_rate": 9.98780370344739e-05, "loss": 2.5625, "step": 2915 }, { "epoch": 0.02, "grad_norm": 1.3204551935195923, "learning_rate": 9.987760356572512e-05, "loss": 2.5974, "step": 2920 }, { "epoch": 0.02, "grad_norm": 1.108158826828003, "learning_rate": 9.987716932899018e-05, "loss": 2.6785, "step": 2925 }, { "epoch": 0.02, "grad_norm": 3.96085524559021, "learning_rate": 9.987673432427576e-05, "loss": 2.5728, "step": 2930 }, { "epoch": 0.02, "grad_norm": 2.139347553253174, "learning_rate": 9.987629855158855e-05, "loss": 2.5864, "step": 2935 }, { "epoch": 0.02, "grad_norm": 2.6015846729278564, "learning_rate": 9.98758620109353e-05, "loss": 2.5421, "step": 2940 }, { "epoch": 0.02, "grad_norm": 0.7267908453941345, "learning_rate": 9.98754247023227e-05, "loss": 2.6279, "step": 2945 }, { "epoch": 0.02, "grad_norm": 1.891913652420044, "learning_rate": 9.987498662575747e-05, "loss": 2.5716, "step": 2950 }, { "epoch": 0.02, "grad_norm": 2.8875906467437744, "learning_rate": 9.987454778124639e-05, "loss": 2.4846, "step": 2955 }, { "epoch": 0.02, "grad_norm": 0.7672701478004456, "learning_rate": 9.987410816879619e-05, "loss": 2.5246, "step": 2960 }, { "epoch": 0.02, "grad_norm": 0.8396023511886597, "learning_rate": 9.987366778841365e-05, "loss": 2.5091, "step": 2965 }, { "epoch": 0.02, "grad_norm": 0.6557952761650085, "learning_rate": 9.987322664010554e-05, "loss": 2.6292, "step": 2970 }, { "epoch": 0.02, "grad_norm": 9.134438514709473, "learning_rate": 9.987278472387867e-05, "loss": 2.5011, "step": 2975 }, { "epoch": 0.02, "grad_norm": 0.8181407451629639, "learning_rate": 9.987234203973986e-05, "loss": 2.6187, "step": 2980 }, { "epoch": 0.02, "grad_norm": 0.7647784352302551, "learning_rate": 9.987189858769586e-05, "loss": 2.5404, "step": 2985 }, { "epoch": 0.02, "grad_norm": 0.9014195203781128, "learning_rate": 9.987145436775357e-05, "loss": 2.5939, "step": 2990 }, { "epoch": 0.02, "grad_norm": 0.83840411901474, "learning_rate": 9.98710093799198e-05, "loss": 2.5809, "step": 2995 }, { "epoch": 0.02, "grad_norm": 0.6832427978515625, "learning_rate": 9.987056362420138e-05, "loss": 2.4703, "step": 3000 }, { "epoch": 0.02, "grad_norm": 0.8244277834892273, "learning_rate": 9.98701171006052e-05, "loss": 2.5107, "step": 3005 }, { "epoch": 0.02, "grad_norm": 1.2349107265472412, "learning_rate": 9.986966980913817e-05, "loss": 2.5897, "step": 3010 }, { "epoch": 0.02, "grad_norm": 0.6188107132911682, "learning_rate": 9.986922174980709e-05, "loss": 2.524, "step": 3015 }, { "epoch": 0.02, "grad_norm": 2.0326879024505615, "learning_rate": 9.986877292261893e-05, "loss": 2.6257, "step": 3020 }, { "epoch": 0.02, "grad_norm": 0.9602454900741577, "learning_rate": 9.986832332758058e-05, "loss": 2.6066, "step": 3025 }, { "epoch": 0.02, "grad_norm": 1.9792391061782837, "learning_rate": 9.986787296469895e-05, "loss": 2.5347, "step": 3030 }, { "epoch": 0.02, "grad_norm": 0.6170005798339844, "learning_rate": 9.986742183398099e-05, "loss": 2.6121, "step": 3035 }, { "epoch": 0.02, "grad_norm": 3.705122470855713, "learning_rate": 9.986696993543366e-05, "loss": 2.5976, "step": 3040 }, { "epoch": 0.02, "grad_norm": 0.8590243458747864, "learning_rate": 9.986651726906387e-05, "loss": 2.6189, "step": 3045 }, { "epoch": 0.02, "grad_norm": 1.8757699728012085, "learning_rate": 9.986606383487862e-05, "loss": 2.4998, "step": 3050 }, { "epoch": 0.02, "grad_norm": 3.960850238800049, "learning_rate": 9.98656096328849e-05, "loss": 2.552, "step": 3055 }, { "epoch": 0.02, "grad_norm": 1.1672649383544922, "learning_rate": 9.98651546630897e-05, "loss": 2.5463, "step": 3060 }, { "epoch": 0.02, "grad_norm": 1.7705445289611816, "learning_rate": 9.986469892550001e-05, "loss": 2.5125, "step": 3065 }, { "epoch": 0.02, "grad_norm": 0.6503201723098755, "learning_rate": 9.986424242012287e-05, "loss": 2.5799, "step": 3070 }, { "epoch": 0.02, "grad_norm": 1.2143852710723877, "learning_rate": 9.986378514696528e-05, "loss": 2.5586, "step": 3075 }, { "epoch": 0.02, "grad_norm": 0.8399056196212769, "learning_rate": 9.98633271060343e-05, "loss": 2.5247, "step": 3080 }, { "epoch": 0.02, "grad_norm": 0.7992956638336182, "learning_rate": 9.9862868297337e-05, "loss": 2.5746, "step": 3085 }, { "epoch": 0.02, "grad_norm": 1.4469751119613647, "learning_rate": 9.98624087208804e-05, "loss": 2.5574, "step": 3090 }, { "epoch": 0.02, "grad_norm": 3.1303155422210693, "learning_rate": 9.98619483766716e-05, "loss": 2.4444, "step": 3095 }, { "epoch": 0.02, "grad_norm": 0.9964423179626465, "learning_rate": 9.98614872647177e-05, "loss": 2.4997, "step": 3100 }, { "epoch": 0.02, "grad_norm": 5.153351306915283, "learning_rate": 9.986102538502579e-05, "loss": 2.4886, "step": 3105 }, { "epoch": 0.02, "grad_norm": 0.9879188537597656, "learning_rate": 9.986056273760297e-05, "loss": 2.6848, "step": 3110 }, { "epoch": 0.02, "grad_norm": 2.624079942703247, "learning_rate": 9.986009932245637e-05, "loss": 2.5125, "step": 3115 }, { "epoch": 0.02, "grad_norm": 1.8067327737808228, "learning_rate": 9.985963513959313e-05, "loss": 2.5102, "step": 3120 }, { "epoch": 0.02, "grad_norm": 0.8255390524864197, "learning_rate": 9.985917018902041e-05, "loss": 2.6584, "step": 3125 }, { "epoch": 0.02, "grad_norm": 0.968376874923706, "learning_rate": 9.985870447074535e-05, "loss": 2.5354, "step": 3130 }, { "epoch": 0.02, "grad_norm": 1.1756937503814697, "learning_rate": 9.985823798477512e-05, "loss": 2.5299, "step": 3135 }, { "epoch": 0.02, "grad_norm": 1.0555535554885864, "learning_rate": 9.985777073111691e-05, "loss": 2.6568, "step": 3140 }, { "epoch": 0.02, "grad_norm": 0.7415378093719482, "learning_rate": 9.985730270977791e-05, "loss": 2.5945, "step": 3145 }, { "epoch": 0.02, "grad_norm": 5.007002830505371, "learning_rate": 9.985683392076534e-05, "loss": 2.5195, "step": 3150 }, { "epoch": 0.02, "grad_norm": 1.8693318367004395, "learning_rate": 9.98563643640864e-05, "loss": 2.5575, "step": 3155 }, { "epoch": 0.02, "grad_norm": 1.9772917032241821, "learning_rate": 9.985589403974833e-05, "loss": 2.6368, "step": 3160 }, { "epoch": 0.02, "grad_norm": 1.1678282022476196, "learning_rate": 9.985542294775838e-05, "loss": 2.5776, "step": 3165 }, { "epoch": 0.03, "grad_norm": 1.3212040662765503, "learning_rate": 9.985495108812377e-05, "loss": 2.5986, "step": 3170 }, { "epoch": 0.03, "grad_norm": 1.1039644479751587, "learning_rate": 9.985447846085181e-05, "loss": 2.4914, "step": 3175 }, { "epoch": 0.03, "grad_norm": 15.303887367248535, "learning_rate": 9.985400506594974e-05, "loss": 2.5741, "step": 3180 }, { "epoch": 0.03, "grad_norm": 1.0935237407684326, "learning_rate": 9.985353090342487e-05, "loss": 2.6182, "step": 3185 }, { "epoch": 0.03, "grad_norm": 0.7300793528556824, "learning_rate": 9.985305597328449e-05, "loss": 2.496, "step": 3190 }, { "epoch": 0.03, "grad_norm": 0.8699349761009216, "learning_rate": 9.985258027553593e-05, "loss": 2.6312, "step": 3195 }, { "epoch": 0.03, "grad_norm": 0.6695219874382019, "learning_rate": 9.98521038101865e-05, "loss": 2.5787, "step": 3200 }, { "epoch": 0.03, "grad_norm": 0.8445362448692322, "learning_rate": 9.985162657724354e-05, "loss": 2.5788, "step": 3205 }, { "epoch": 0.03, "grad_norm": 0.9595109820365906, "learning_rate": 9.98511485767144e-05, "loss": 2.6037, "step": 3210 }, { "epoch": 0.03, "grad_norm": 0.7013242244720459, "learning_rate": 9.985066980860642e-05, "loss": 2.6216, "step": 3215 }, { "epoch": 0.03, "grad_norm": 0.8572027087211609, "learning_rate": 9.9850190272927e-05, "loss": 2.513, "step": 3220 }, { "epoch": 0.03, "grad_norm": 1.1105091571807861, "learning_rate": 9.984970996968352e-05, "loss": 2.5318, "step": 3225 }, { "epoch": 0.03, "grad_norm": 4.040609359741211, "learning_rate": 9.984922889888335e-05, "loss": 2.6298, "step": 3230 }, { "epoch": 0.03, "grad_norm": 1.677856206893921, "learning_rate": 9.984874706053393e-05, "loss": 2.534, "step": 3235 }, { "epoch": 0.03, "grad_norm": 0.8891656994819641, "learning_rate": 9.984826445464266e-05, "loss": 2.5702, "step": 3240 }, { "epoch": 0.03, "grad_norm": 0.8716422915458679, "learning_rate": 9.984778108121698e-05, "loss": 2.5155, "step": 3245 }, { "epoch": 0.03, "grad_norm": 0.7925968170166016, "learning_rate": 9.984729694026431e-05, "loss": 2.627, "step": 3250 }, { "epoch": 0.03, "grad_norm": 0.991220235824585, "learning_rate": 9.984681203179214e-05, "loss": 2.5597, "step": 3255 }, { "epoch": 0.03, "grad_norm": 1.3146400451660156, "learning_rate": 9.984632635580792e-05, "loss": 2.5577, "step": 3260 }, { "epoch": 0.03, "grad_norm": 9.56419563293457, "learning_rate": 9.984583991231913e-05, "loss": 2.5954, "step": 3265 }, { "epoch": 0.03, "grad_norm": 0.9446741938591003, "learning_rate": 9.984535270133324e-05, "loss": 2.5556, "step": 3270 }, { "epoch": 0.03, "grad_norm": 0.8494237661361694, "learning_rate": 9.984486472285777e-05, "loss": 2.5474, "step": 3275 }, { "epoch": 0.03, "grad_norm": 0.7279638648033142, "learning_rate": 9.984437597690024e-05, "loss": 2.5913, "step": 3280 }, { "epoch": 0.03, "grad_norm": 1.0752673149108887, "learning_rate": 9.984388646346817e-05, "loss": 2.5335, "step": 3285 }, { "epoch": 0.03, "grad_norm": 1.5325186252593994, "learning_rate": 9.984339618256909e-05, "loss": 2.5278, "step": 3290 }, { "epoch": 0.03, "grad_norm": 11.744277954101562, "learning_rate": 9.984290513421057e-05, "loss": 2.6065, "step": 3295 }, { "epoch": 0.03, "grad_norm": 2.1811797618865967, "learning_rate": 9.984241331840015e-05, "loss": 2.5716, "step": 3300 }, { "epoch": 0.03, "grad_norm": 1.5920522212982178, "learning_rate": 9.984192073514538e-05, "loss": 2.5172, "step": 3305 }, { "epoch": 0.03, "grad_norm": 26.787242889404297, "learning_rate": 9.98414273844539e-05, "loss": 2.6891, "step": 3310 }, { "epoch": 0.03, "grad_norm": 0.9135124087333679, "learning_rate": 9.984093326633325e-05, "loss": 2.6022, "step": 3315 }, { "epoch": 0.03, "grad_norm": 0.8413210511207581, "learning_rate": 9.98404383807911e-05, "loss": 2.5297, "step": 3320 }, { "epoch": 0.03, "grad_norm": 0.7774330973625183, "learning_rate": 9.983994272783502e-05, "loss": 2.5419, "step": 3325 }, { "epoch": 0.03, "grad_norm": 1.168771505355835, "learning_rate": 9.983944630747266e-05, "loss": 2.5547, "step": 3330 }, { "epoch": 0.03, "grad_norm": 0.8908506631851196, "learning_rate": 9.983894911971167e-05, "loss": 2.5294, "step": 3335 }, { "epoch": 0.03, "grad_norm": 0.8098046183586121, "learning_rate": 9.983845116455969e-05, "loss": 2.5833, "step": 3340 }, { "epoch": 0.03, "grad_norm": 10.539358139038086, "learning_rate": 9.983795244202439e-05, "loss": 2.5607, "step": 3345 }, { "epoch": 0.03, "grad_norm": 2.144416332244873, "learning_rate": 9.983745295211347e-05, "loss": 2.5353, "step": 3350 }, { "epoch": 0.03, "grad_norm": 1.8186370134353638, "learning_rate": 9.983695269483459e-05, "loss": 2.5513, "step": 3355 }, { "epoch": 0.03, "grad_norm": 17.618772506713867, "learning_rate": 9.983645167019546e-05, "loss": 2.5684, "step": 3360 }, { "epoch": 0.03, "grad_norm": 1.4243966341018677, "learning_rate": 9.983594987820381e-05, "loss": 2.5921, "step": 3365 }, { "epoch": 0.03, "grad_norm": 2.2420129776000977, "learning_rate": 9.983544731886735e-05, "loss": 2.5395, "step": 3370 }, { "epoch": 0.03, "grad_norm": 1.0887452363967896, "learning_rate": 9.983494399219386e-05, "loss": 2.5306, "step": 3375 }, { "epoch": 0.03, "grad_norm": 1.1642740964889526, "learning_rate": 9.983443989819103e-05, "loss": 2.5439, "step": 3380 }, { "epoch": 0.03, "grad_norm": 3.7865490913391113, "learning_rate": 9.983393503686665e-05, "loss": 2.4518, "step": 3385 }, { "epoch": 0.03, "grad_norm": 0.8654516339302063, "learning_rate": 9.983342940822849e-05, "loss": 2.4727, "step": 3390 }, { "epoch": 0.03, "grad_norm": 1.2857521772384644, "learning_rate": 9.983292301228432e-05, "loss": 2.5028, "step": 3395 }, { "epoch": 0.03, "grad_norm": 0.975713849067688, "learning_rate": 9.983241584904197e-05, "loss": 2.4435, "step": 3400 }, { "epoch": 0.03, "grad_norm": 1.524484634399414, "learning_rate": 9.983190791850925e-05, "loss": 2.5062, "step": 3405 }, { "epoch": 0.03, "grad_norm": 1.7540284395217896, "learning_rate": 9.983139922069394e-05, "loss": 2.6116, "step": 3410 }, { "epoch": 0.03, "grad_norm": 0.6647650003433228, "learning_rate": 9.983088975560391e-05, "loss": 2.5257, "step": 3415 }, { "epoch": 0.03, "grad_norm": 1.9461784362792969, "learning_rate": 9.983037952324699e-05, "loss": 2.4752, "step": 3420 }, { "epoch": 0.03, "grad_norm": 0.8311464190483093, "learning_rate": 9.982986852363102e-05, "loss": 2.5565, "step": 3425 }, { "epoch": 0.03, "grad_norm": 1.2381539344787598, "learning_rate": 9.98293567567639e-05, "loss": 2.4489, "step": 3430 }, { "epoch": 0.03, "grad_norm": 0.7437267303466797, "learning_rate": 9.982884422265349e-05, "loss": 2.5125, "step": 3435 }, { "epoch": 0.03, "grad_norm": 2.480928897857666, "learning_rate": 9.982833092130768e-05, "loss": 2.6542, "step": 3440 }, { "epoch": 0.03, "grad_norm": 1.302895426750183, "learning_rate": 9.982781685273438e-05, "loss": 2.4461, "step": 3445 }, { "epoch": 0.03, "grad_norm": 0.6435883641242981, "learning_rate": 9.982730201694153e-05, "loss": 2.5266, "step": 3450 }, { "epoch": 0.03, "grad_norm": 0.6683421730995178, "learning_rate": 9.982678641393701e-05, "loss": 2.5216, "step": 3455 }, { "epoch": 0.03, "grad_norm": 1.317556619644165, "learning_rate": 9.98262700437288e-05, "loss": 2.5438, "step": 3460 }, { "epoch": 0.03, "grad_norm": 2.4007017612457275, "learning_rate": 9.982575290632481e-05, "loss": 2.4536, "step": 3465 }, { "epoch": 0.03, "grad_norm": 1.4988950490951538, "learning_rate": 9.982523500173304e-05, "loss": 2.5569, "step": 3470 }, { "epoch": 0.03, "grad_norm": 1.0140033960342407, "learning_rate": 9.982471632996144e-05, "loss": 2.457, "step": 3475 }, { "epoch": 0.03, "grad_norm": 1.6312874555587769, "learning_rate": 9.9824196891018e-05, "loss": 2.6171, "step": 3480 }, { "epoch": 0.03, "grad_norm": 1.7508939504623413, "learning_rate": 9.982367668491075e-05, "loss": 2.5587, "step": 3485 }, { "epoch": 0.03, "grad_norm": 1.597745656967163, "learning_rate": 9.982315571164765e-05, "loss": 2.5004, "step": 3490 }, { "epoch": 0.03, "grad_norm": 0.9887745380401611, "learning_rate": 9.982263397123675e-05, "loss": 2.5569, "step": 3495 }, { "epoch": 0.03, "grad_norm": 1.2374274730682373, "learning_rate": 9.982211146368608e-05, "loss": 2.6261, "step": 3500 }, { "epoch": 0.03, "grad_norm": 1.375279188156128, "learning_rate": 9.982158818900369e-05, "loss": 2.4822, "step": 3505 }, { "epoch": 0.03, "grad_norm": 0.6986528635025024, "learning_rate": 9.982106414719762e-05, "loss": 2.4449, "step": 3510 }, { "epoch": 0.03, "grad_norm": 1.383217215538025, "learning_rate": 9.982053933827595e-05, "loss": 2.6297, "step": 3515 }, { "epoch": 0.03, "grad_norm": 0.8467180728912354, "learning_rate": 9.982001376224677e-05, "loss": 2.5074, "step": 3520 }, { "epoch": 0.03, "grad_norm": 2.666476011276245, "learning_rate": 9.981948741911816e-05, "loss": 2.6461, "step": 3525 }, { "epoch": 0.03, "grad_norm": 0.8575208783149719, "learning_rate": 9.981896030889821e-05, "loss": 2.5155, "step": 3530 }, { "epoch": 0.03, "grad_norm": 0.6981438994407654, "learning_rate": 9.981843243159507e-05, "loss": 2.5504, "step": 3535 }, { "epoch": 0.03, "grad_norm": 1.2238153219223022, "learning_rate": 9.981790378721684e-05, "loss": 2.5114, "step": 3540 }, { "epoch": 0.03, "grad_norm": 0.9935281872749329, "learning_rate": 9.98173743757717e-05, "loss": 2.5118, "step": 3545 }, { "epoch": 0.03, "grad_norm": 2.5141870975494385, "learning_rate": 9.981684419726775e-05, "loss": 2.6055, "step": 3550 }, { "epoch": 0.03, "grad_norm": 6.857272148132324, "learning_rate": 9.981631325171318e-05, "loss": 2.4613, "step": 3555 }, { "epoch": 0.03, "grad_norm": 1.226979374885559, "learning_rate": 9.981578153911616e-05, "loss": 2.4915, "step": 3560 }, { "epoch": 0.03, "grad_norm": 3.7647297382354736, "learning_rate": 9.981524905948487e-05, "loss": 2.5526, "step": 3565 }, { "epoch": 0.03, "grad_norm": 0.9020975828170776, "learning_rate": 9.981471581282753e-05, "loss": 2.5152, "step": 3570 }, { "epoch": 0.03, "grad_norm": 4.5760626792907715, "learning_rate": 9.981418179915234e-05, "loss": 2.4919, "step": 3575 }, { "epoch": 0.03, "grad_norm": 1.1118589639663696, "learning_rate": 9.981364701846752e-05, "loss": 2.4712, "step": 3580 }, { "epoch": 0.03, "grad_norm": 1.4426705837249756, "learning_rate": 9.981311147078128e-05, "loss": 2.4916, "step": 3585 }, { "epoch": 0.03, "grad_norm": 9.646313667297363, "learning_rate": 9.981257515610192e-05, "loss": 2.465, "step": 3590 }, { "epoch": 0.03, "grad_norm": 0.7251165509223938, "learning_rate": 9.981203807443764e-05, "loss": 2.4943, "step": 3595 }, { "epoch": 0.03, "grad_norm": 0.698371946811676, "learning_rate": 9.981150022579676e-05, "loss": 2.5477, "step": 3600 }, { "epoch": 0.03, "grad_norm": 2.0286972522735596, "learning_rate": 9.981096161018754e-05, "loss": 2.5409, "step": 3605 }, { "epoch": 0.03, "grad_norm": 0.8140892386436462, "learning_rate": 9.981042222761826e-05, "loss": 2.5126, "step": 3610 }, { "epoch": 0.03, "grad_norm": 0.8781805038452148, "learning_rate": 9.980988207809726e-05, "loss": 2.4917, "step": 3615 }, { "epoch": 0.03, "grad_norm": 1.1943638324737549, "learning_rate": 9.98093411616328e-05, "loss": 2.5264, "step": 3620 }, { "epoch": 0.03, "grad_norm": 0.8444727063179016, "learning_rate": 9.980879947823327e-05, "loss": 2.7064, "step": 3625 }, { "epoch": 0.03, "grad_norm": 1.8897470235824585, "learning_rate": 9.980825702790696e-05, "loss": 2.429, "step": 3630 }, { "epoch": 0.03, "grad_norm": 0.8530240058898926, "learning_rate": 9.980771381066227e-05, "loss": 2.5633, "step": 3635 }, { "epoch": 0.03, "grad_norm": 1.019193410873413, "learning_rate": 9.980716982650753e-05, "loss": 2.5197, "step": 3640 }, { "epoch": 0.03, "grad_norm": 1.5260518789291382, "learning_rate": 9.980662507545113e-05, "loss": 2.6042, "step": 3645 }, { "epoch": 0.03, "grad_norm": 0.9051507115364075, "learning_rate": 9.980607955750146e-05, "loss": 2.4934, "step": 3650 }, { "epoch": 0.03, "grad_norm": 1.662142038345337, "learning_rate": 9.98055332726669e-05, "loss": 2.6125, "step": 3655 }, { "epoch": 0.03, "grad_norm": 1.2840546369552612, "learning_rate": 9.980498622095588e-05, "loss": 2.4327, "step": 3660 }, { "epoch": 0.03, "grad_norm": 1.15514075756073, "learning_rate": 9.980443840237681e-05, "loss": 2.4918, "step": 3665 }, { "epoch": 0.03, "grad_norm": 0.8532174229621887, "learning_rate": 9.980388981693816e-05, "loss": 2.6136, "step": 3670 }, { "epoch": 0.03, "grad_norm": 1.4051742553710938, "learning_rate": 9.980334046464832e-05, "loss": 2.6388, "step": 3675 }, { "epoch": 0.03, "grad_norm": 0.7405477166175842, "learning_rate": 9.980279034551579e-05, "loss": 2.6699, "step": 3680 }, { "epoch": 0.03, "grad_norm": 0.8891518712043762, "learning_rate": 9.980223945954902e-05, "loss": 2.5558, "step": 3685 }, { "epoch": 0.03, "grad_norm": 0.7266207337379456, "learning_rate": 9.980168780675651e-05, "loss": 2.4098, "step": 3690 }, { "epoch": 0.03, "grad_norm": 0.9755536317825317, "learning_rate": 9.980113538714675e-05, "loss": 2.5315, "step": 3695 }, { "epoch": 0.03, "grad_norm": 16.053264617919922, "learning_rate": 9.980058220072825e-05, "loss": 2.5714, "step": 3700 }, { "epoch": 0.03, "grad_norm": 0.9983071088790894, "learning_rate": 9.98000282475095e-05, "loss": 2.5119, "step": 3705 }, { "epoch": 0.03, "grad_norm": 28.287281036376953, "learning_rate": 9.979947352749906e-05, "loss": 2.5175, "step": 3710 }, { "epoch": 0.03, "grad_norm": 1.246410846710205, "learning_rate": 9.979891804070545e-05, "loss": 2.5181, "step": 3715 }, { "epoch": 0.03, "grad_norm": 1.5269850492477417, "learning_rate": 9.979836178713724e-05, "loss": 2.592, "step": 3720 }, { "epoch": 0.03, "grad_norm": 1.2013732194900513, "learning_rate": 9.979780476680298e-05, "loss": 2.5303, "step": 3725 }, { "epoch": 0.03, "grad_norm": 0.9525001645088196, "learning_rate": 9.979724697971125e-05, "loss": 2.4659, "step": 3730 }, { "epoch": 0.03, "grad_norm": 1.543030858039856, "learning_rate": 9.979668842587065e-05, "loss": 2.585, "step": 3735 }, { "epoch": 0.03, "grad_norm": 10.039307594299316, "learning_rate": 9.979612910528976e-05, "loss": 2.5025, "step": 3740 }, { "epoch": 0.03, "grad_norm": 0.7964401841163635, "learning_rate": 9.979556901797723e-05, "loss": 2.6029, "step": 3745 }, { "epoch": 0.03, "grad_norm": 0.7487723231315613, "learning_rate": 9.979500816394165e-05, "loss": 2.4487, "step": 3750 }, { "epoch": 0.03, "grad_norm": 1.6278318166732788, "learning_rate": 9.979444654319165e-05, "loss": 2.5177, "step": 3755 }, { "epoch": 0.03, "grad_norm": 1.0509902238845825, "learning_rate": 9.979388415573591e-05, "loss": 2.4638, "step": 3760 }, { "epoch": 0.03, "grad_norm": 0.920802116394043, "learning_rate": 9.979332100158307e-05, "loss": 2.5439, "step": 3765 }, { "epoch": 0.03, "grad_norm": 0.9470750689506531, "learning_rate": 9.979275708074179e-05, "loss": 2.4859, "step": 3770 }, { "epoch": 0.03, "grad_norm": 1.3247926235198975, "learning_rate": 9.979219239322078e-05, "loss": 2.563, "step": 3775 }, { "epoch": 0.03, "grad_norm": 0.8123425841331482, "learning_rate": 9.979162693902871e-05, "loss": 2.4905, "step": 3780 }, { "epoch": 0.03, "grad_norm": 0.5892263054847717, "learning_rate": 9.97910607181743e-05, "loss": 2.5201, "step": 3785 }, { "epoch": 0.03, "grad_norm": 0.6774473190307617, "learning_rate": 9.979049373066626e-05, "loss": 2.4716, "step": 3790 }, { "epoch": 0.03, "grad_norm": 0.9502044320106506, "learning_rate": 9.978992597651334e-05, "loss": 2.5116, "step": 3795 }, { "epoch": 0.03, "grad_norm": 18.373802185058594, "learning_rate": 9.978935745572424e-05, "loss": 2.5238, "step": 3800 }, { "epoch": 0.03, "grad_norm": 1.30051851272583, "learning_rate": 9.978878816830776e-05, "loss": 2.4439, "step": 3805 }, { "epoch": 0.03, "grad_norm": 0.6439379453659058, "learning_rate": 9.978821811427263e-05, "loss": 2.4787, "step": 3810 }, { "epoch": 0.03, "grad_norm": 0.7634502053260803, "learning_rate": 9.978764729362766e-05, "loss": 2.5454, "step": 3815 }, { "epoch": 0.03, "grad_norm": 0.7452027201652527, "learning_rate": 9.978707570638162e-05, "loss": 2.4662, "step": 3820 }, { "epoch": 0.03, "grad_norm": 9.542091369628906, "learning_rate": 9.978650335254332e-05, "loss": 2.4785, "step": 3825 }, { "epoch": 0.03, "grad_norm": 1.1908836364746094, "learning_rate": 9.978593023212155e-05, "loss": 2.5736, "step": 3830 }, { "epoch": 0.03, "grad_norm": 1.947479248046875, "learning_rate": 9.978535634512515e-05, "loss": 2.4677, "step": 3835 }, { "epoch": 0.03, "grad_norm": 0.9893966317176819, "learning_rate": 9.978478169156298e-05, "loss": 2.4952, "step": 3840 }, { "epoch": 0.03, "grad_norm": 0.8016408085823059, "learning_rate": 9.978420627144385e-05, "loss": 2.5454, "step": 3845 }, { "epoch": 0.03, "grad_norm": 0.702154815196991, "learning_rate": 9.978363008477665e-05, "loss": 2.4459, "step": 3850 }, { "epoch": 0.03, "grad_norm": 0.6297614574432373, "learning_rate": 9.978305313157022e-05, "loss": 2.528, "step": 3855 }, { "epoch": 0.03, "grad_norm": 1.0170215368270874, "learning_rate": 9.978247541183348e-05, "loss": 2.5385, "step": 3860 }, { "epoch": 0.03, "grad_norm": 1.7152056694030762, "learning_rate": 9.978189692557529e-05, "loss": 2.5408, "step": 3865 }, { "epoch": 0.03, "grad_norm": 0.9676519632339478, "learning_rate": 9.978131767280456e-05, "loss": 2.5273, "step": 3870 }, { "epoch": 0.03, "grad_norm": 3.040776252746582, "learning_rate": 9.978073765353026e-05, "loss": 2.3205, "step": 3875 }, { "epoch": 0.03, "grad_norm": 1.9810748100280762, "learning_rate": 9.978015686776125e-05, "loss": 2.5646, "step": 3880 }, { "epoch": 0.03, "grad_norm": 2.523010015487671, "learning_rate": 9.977957531550652e-05, "loss": 2.5544, "step": 3885 }, { "epoch": 0.03, "grad_norm": 0.9049906134605408, "learning_rate": 9.977899299677502e-05, "loss": 2.5289, "step": 3890 }, { "epoch": 0.03, "grad_norm": 1.0273253917694092, "learning_rate": 9.97784099115757e-05, "loss": 2.5007, "step": 3895 }, { "epoch": 0.03, "grad_norm": 1.2546312808990479, "learning_rate": 9.977782605991754e-05, "loss": 2.5753, "step": 3900 }, { "epoch": 0.03, "grad_norm": 0.7546401619911194, "learning_rate": 9.977724144180953e-05, "loss": 2.5046, "step": 3905 }, { "epoch": 0.03, "grad_norm": 0.946729302406311, "learning_rate": 9.977665605726067e-05, "loss": 2.5218, "step": 3910 }, { "epoch": 0.03, "grad_norm": 1.0106152296066284, "learning_rate": 9.977606990628001e-05, "loss": 2.4549, "step": 3915 }, { "epoch": 0.03, "grad_norm": 1.141021966934204, "learning_rate": 9.977548298887651e-05, "loss": 2.5714, "step": 3920 }, { "epoch": 0.03, "grad_norm": 3.473891496658325, "learning_rate": 9.977489530505927e-05, "loss": 2.4456, "step": 3925 }, { "epoch": 0.03, "grad_norm": 1.0176597833633423, "learning_rate": 9.97743068548373e-05, "loss": 2.4192, "step": 3930 }, { "epoch": 0.03, "grad_norm": 0.8160073757171631, "learning_rate": 9.977371763821966e-05, "loss": 2.5231, "step": 3935 }, { "epoch": 0.03, "grad_norm": 9.338330268859863, "learning_rate": 9.977312765521545e-05, "loss": 2.507, "step": 3940 }, { "epoch": 0.03, "grad_norm": 0.7023669481277466, "learning_rate": 9.977253690583372e-05, "loss": 2.5117, "step": 3945 }, { "epoch": 0.03, "grad_norm": 0.7150111794471741, "learning_rate": 9.97719453900836e-05, "loss": 2.4741, "step": 3950 }, { "epoch": 0.03, "grad_norm": 1.5828582048416138, "learning_rate": 9.977135310797418e-05, "loss": 2.4403, "step": 3955 }, { "epoch": 0.03, "grad_norm": 16.272518157958984, "learning_rate": 9.977076005951458e-05, "loss": 2.4874, "step": 3960 }, { "epoch": 0.03, "grad_norm": 1.3931905031204224, "learning_rate": 9.977016624471392e-05, "loss": 2.5574, "step": 3965 }, { "epoch": 0.03, "grad_norm": 0.9250815510749817, "learning_rate": 9.976957166358137e-05, "loss": 2.3903, "step": 3970 }, { "epoch": 0.03, "grad_norm": 1.663214087486267, "learning_rate": 9.976897631612607e-05, "loss": 2.4318, "step": 3975 }, { "epoch": 0.03, "grad_norm": 0.9962267279624939, "learning_rate": 9.97683802023572e-05, "loss": 2.5191, "step": 3980 }, { "epoch": 0.03, "grad_norm": 0.7311832904815674, "learning_rate": 9.976778332228389e-05, "loss": 2.5336, "step": 3985 }, { "epoch": 0.03, "grad_norm": 0.6227299571037292, "learning_rate": 9.97671856759154e-05, "loss": 2.4366, "step": 3990 }, { "epoch": 0.03, "grad_norm": 0.7928274869918823, "learning_rate": 9.976658726326088e-05, "loss": 2.525, "step": 3995 }, { "epoch": 0.03, "grad_norm": 9.608405113220215, "learning_rate": 9.976598808432957e-05, "loss": 2.5196, "step": 4000 }, { "epoch": 0.03, "grad_norm": 1.321887731552124, "learning_rate": 9.976538813913068e-05, "loss": 2.4798, "step": 4005 }, { "epoch": 0.03, "grad_norm": 19.59874725341797, "learning_rate": 9.976478742767347e-05, "loss": 2.5865, "step": 4010 }, { "epoch": 0.03, "grad_norm": 1.4570783376693726, "learning_rate": 9.976418594996716e-05, "loss": 2.4878, "step": 4015 }, { "epoch": 0.03, "grad_norm": 0.8006621599197388, "learning_rate": 9.976358370602103e-05, "loss": 2.4905, "step": 4020 }, { "epoch": 0.03, "grad_norm": 11.88657283782959, "learning_rate": 9.976298069584438e-05, "loss": 2.503, "step": 4025 }, { "epoch": 0.03, "grad_norm": 0.9018133282661438, "learning_rate": 9.976237691944643e-05, "loss": 2.4983, "step": 4030 }, { "epoch": 0.03, "grad_norm": 3.900397777557373, "learning_rate": 9.976177237683653e-05, "loss": 2.6024, "step": 4035 }, { "epoch": 0.03, "grad_norm": 3.2990190982818604, "learning_rate": 9.976116706802395e-05, "loss": 2.4815, "step": 4040 }, { "epoch": 0.03, "grad_norm": 1.352231740951538, "learning_rate": 9.976056099301805e-05, "loss": 2.5453, "step": 4045 }, { "epoch": 0.03, "grad_norm": 1.1283690929412842, "learning_rate": 9.975995415182812e-05, "loss": 2.592, "step": 4050 }, { "epoch": 0.03, "grad_norm": 0.9985553026199341, "learning_rate": 9.975934654446355e-05, "loss": 2.4858, "step": 4055 }, { "epoch": 0.03, "grad_norm": 1.4596166610717773, "learning_rate": 9.975873817093367e-05, "loss": 2.5337, "step": 4060 }, { "epoch": 0.03, "grad_norm": 0.814937174320221, "learning_rate": 9.975812903124782e-05, "loss": 2.5066, "step": 4065 }, { "epoch": 0.03, "grad_norm": 15.678441047668457, "learning_rate": 9.975751912541544e-05, "loss": 2.4737, "step": 4070 }, { "epoch": 0.03, "grad_norm": 0.8281909227371216, "learning_rate": 9.975690845344588e-05, "loss": 2.4883, "step": 4075 }, { "epoch": 0.03, "grad_norm": 4.475855350494385, "learning_rate": 9.975629701534854e-05, "loss": 2.5173, "step": 4080 }, { "epoch": 0.03, "grad_norm": 0.8197420835494995, "learning_rate": 9.975568481113285e-05, "loss": 2.5011, "step": 4085 }, { "epoch": 0.03, "grad_norm": 6.603283405303955, "learning_rate": 9.975507184080823e-05, "loss": 2.5188, "step": 4090 }, { "epoch": 0.03, "grad_norm": 0.743004322052002, "learning_rate": 9.975445810438413e-05, "loss": 2.5169, "step": 4095 }, { "epoch": 0.03, "grad_norm": 0.8768494129180908, "learning_rate": 9.975384360186999e-05, "loss": 2.4576, "step": 4100 }, { "epoch": 0.03, "grad_norm": 1.1234978437423706, "learning_rate": 9.975322833327527e-05, "loss": 2.4859, "step": 4105 }, { "epoch": 0.03, "grad_norm": 22.23385238647461, "learning_rate": 9.975261229860944e-05, "loss": 2.5377, "step": 4110 }, { "epoch": 0.03, "grad_norm": 0.8795993328094482, "learning_rate": 9.975199549788199e-05, "loss": 2.4916, "step": 4115 }, { "epoch": 0.03, "grad_norm": 2.495150089263916, "learning_rate": 9.975137793110243e-05, "loss": 2.5783, "step": 4120 }, { "epoch": 0.03, "grad_norm": 0.9077249765396118, "learning_rate": 9.975075959828026e-05, "loss": 2.4693, "step": 4125 }, { "epoch": 0.03, "grad_norm": 4.634250164031982, "learning_rate": 9.975014049942499e-05, "loss": 2.4894, "step": 4130 }, { "epoch": 0.03, "grad_norm": 1.2079253196716309, "learning_rate": 9.974952063454615e-05, "loss": 2.5075, "step": 4135 }, { "epoch": 0.03, "grad_norm": 1.4984222650527954, "learning_rate": 9.974890000365332e-05, "loss": 2.5434, "step": 4140 }, { "epoch": 0.03, "grad_norm": 0.6156737208366394, "learning_rate": 9.9748278606756e-05, "loss": 2.4966, "step": 4145 }, { "epoch": 0.03, "grad_norm": 0.9832416772842407, "learning_rate": 9.974765644386382e-05, "loss": 2.4411, "step": 4150 }, { "epoch": 0.03, "grad_norm": 0.8693953156471252, "learning_rate": 9.974703351498631e-05, "loss": 2.5513, "step": 4155 }, { "epoch": 0.03, "grad_norm": 1.6250112056732178, "learning_rate": 9.974640982013307e-05, "loss": 2.4188, "step": 4160 }, { "epoch": 0.03, "grad_norm": 0.9275277256965637, "learning_rate": 9.974578535931374e-05, "loss": 2.4755, "step": 4165 }, { "epoch": 0.03, "grad_norm": 2.183382749557495, "learning_rate": 9.974516013253789e-05, "loss": 2.486, "step": 4170 }, { "epoch": 0.03, "grad_norm": 2.6413698196411133, "learning_rate": 9.974453413981517e-05, "loss": 2.4407, "step": 4175 }, { "epoch": 0.03, "grad_norm": 3.2773215770721436, "learning_rate": 9.974390738115521e-05, "loss": 2.6026, "step": 4180 }, { "epoch": 0.03, "grad_norm": 5.735229969024658, "learning_rate": 9.974327985656767e-05, "loss": 2.5439, "step": 4185 }, { "epoch": 0.03, "grad_norm": 0.7258651256561279, "learning_rate": 9.974265156606221e-05, "loss": 2.5356, "step": 4190 }, { "epoch": 0.03, "grad_norm": 0.6889208555221558, "learning_rate": 9.974202250964849e-05, "loss": 2.4688, "step": 4195 }, { "epoch": 0.03, "grad_norm": 8.396142959594727, "learning_rate": 9.974139268733623e-05, "loss": 2.5996, "step": 4200 }, { "epoch": 0.03, "grad_norm": 0.9942218661308289, "learning_rate": 9.974076209913508e-05, "loss": 2.4998, "step": 4205 }, { "epoch": 0.03, "grad_norm": 1.3743064403533936, "learning_rate": 9.974013074505477e-05, "loss": 2.5067, "step": 4210 }, { "epoch": 0.03, "grad_norm": 0.8313949108123779, "learning_rate": 9.973949862510503e-05, "loss": 2.537, "step": 4215 }, { "epoch": 0.03, "grad_norm": 0.6077825427055359, "learning_rate": 9.97388657392956e-05, "loss": 2.5661, "step": 4220 }, { "epoch": 0.03, "grad_norm": 12.75197982788086, "learning_rate": 9.97382320876362e-05, "loss": 2.4197, "step": 4225 }, { "epoch": 0.03, "grad_norm": 0.7538285851478577, "learning_rate": 9.973759767013662e-05, "loss": 2.5375, "step": 4230 }, { "epoch": 0.03, "grad_norm": 7.1428656578063965, "learning_rate": 9.973696248680659e-05, "loss": 2.5785, "step": 4235 }, { "epoch": 0.03, "grad_norm": 0.7587202787399292, "learning_rate": 9.97363265376559e-05, "loss": 2.4588, "step": 4240 }, { "epoch": 0.03, "grad_norm": 0.9672504663467407, "learning_rate": 9.973568982269436e-05, "loss": 2.4673, "step": 4245 }, { "epoch": 0.03, "grad_norm": 24.25544548034668, "learning_rate": 9.973505234193177e-05, "loss": 2.4604, "step": 4250 }, { "epoch": 0.03, "grad_norm": 1.593388557434082, "learning_rate": 9.973441409537793e-05, "loss": 2.4627, "step": 4255 }, { "epoch": 0.03, "grad_norm": 0.7687398791313171, "learning_rate": 9.973377508304269e-05, "loss": 2.4366, "step": 4260 }, { "epoch": 0.03, "grad_norm": 7.224075794219971, "learning_rate": 9.973313530493586e-05, "loss": 2.5037, "step": 4265 }, { "epoch": 0.03, "grad_norm": 0.9114329814910889, "learning_rate": 9.973249476106732e-05, "loss": 2.5985, "step": 4270 }, { "epoch": 0.03, "grad_norm": 0.9974742531776428, "learning_rate": 9.97318534514469e-05, "loss": 2.5112, "step": 4275 }, { "epoch": 0.03, "grad_norm": 0.8431203961372375, "learning_rate": 9.97312113760845e-05, "loss": 2.4326, "step": 4280 }, { "epoch": 0.03, "grad_norm": 5.112524032592773, "learning_rate": 9.973056853499001e-05, "loss": 2.4111, "step": 4285 }, { "epoch": 0.03, "grad_norm": 1.069179654121399, "learning_rate": 9.972992492817331e-05, "loss": 2.4831, "step": 4290 }, { "epoch": 0.03, "grad_norm": 0.8051745891571045, "learning_rate": 9.972928055564432e-05, "loss": 2.5634, "step": 4295 }, { "epoch": 0.03, "grad_norm": 3.5543534755706787, "learning_rate": 9.972863541741297e-05, "loss": 2.5078, "step": 4300 }, { "epoch": 0.03, "grad_norm": 1.8455121517181396, "learning_rate": 9.972798951348918e-05, "loss": 2.509, "step": 4305 }, { "epoch": 0.03, "grad_norm": 1.1424040794372559, "learning_rate": 9.972734284388288e-05, "loss": 2.5366, "step": 4310 }, { "epoch": 0.03, "grad_norm": 1.2641465663909912, "learning_rate": 9.972669540860406e-05, "loss": 2.489, "step": 4315 }, { "epoch": 0.03, "grad_norm": 1.1265602111816406, "learning_rate": 9.972604720766268e-05, "loss": 2.5694, "step": 4320 }, { "epoch": 0.03, "grad_norm": 1.804335117340088, "learning_rate": 9.97253982410687e-05, "loss": 2.563, "step": 4325 }, { "epoch": 0.03, "grad_norm": 2.18552303314209, "learning_rate": 9.972474850883213e-05, "loss": 2.5735, "step": 4330 }, { "epoch": 0.03, "grad_norm": 0.739506185054779, "learning_rate": 9.972409801096299e-05, "loss": 2.5111, "step": 4335 }, { "epoch": 0.03, "grad_norm": 0.8701598048210144, "learning_rate": 9.972344674747125e-05, "loss": 2.4609, "step": 4340 }, { "epoch": 0.03, "grad_norm": 0.996677577495575, "learning_rate": 9.972279471836698e-05, "loss": 2.4808, "step": 4345 }, { "epoch": 0.03, "grad_norm": 3.2452163696289062, "learning_rate": 9.972214192366021e-05, "loss": 2.4535, "step": 4350 }, { "epoch": 0.03, "grad_norm": 1.0309845209121704, "learning_rate": 9.972148836336098e-05, "loss": 2.4913, "step": 4355 }, { "epoch": 0.03, "grad_norm": 1.341454267501831, "learning_rate": 9.972083403747937e-05, "loss": 2.4594, "step": 4360 }, { "epoch": 0.03, "grad_norm": 0.8861576318740845, "learning_rate": 9.972017894602542e-05, "loss": 2.4941, "step": 4365 }, { "epoch": 0.03, "grad_norm": 1.1018404960632324, "learning_rate": 9.971952308900925e-05, "loss": 2.4665, "step": 4370 }, { "epoch": 0.03, "grad_norm": 0.5916318297386169, "learning_rate": 9.971886646644096e-05, "loss": 2.4398, "step": 4375 }, { "epoch": 0.03, "grad_norm": 0.5977169275283813, "learning_rate": 9.971820907833065e-05, "loss": 2.4232, "step": 4380 }, { "epoch": 0.03, "grad_norm": 11.125341415405273, "learning_rate": 9.971755092468843e-05, "loss": 2.4987, "step": 4385 }, { "epoch": 0.03, "grad_norm": 0.6846534609794617, "learning_rate": 9.971689200552445e-05, "loss": 2.4739, "step": 4390 }, { "epoch": 0.03, "grad_norm": 0.6106178164482117, "learning_rate": 9.971623232084886e-05, "loss": 2.4312, "step": 4395 }, { "epoch": 0.03, "grad_norm": 0.8904944658279419, "learning_rate": 9.971557187067179e-05, "loss": 2.5104, "step": 4400 }, { "epoch": 0.03, "grad_norm": 0.7764661312103271, "learning_rate": 9.971491065500345e-05, "loss": 2.4847, "step": 4405 }, { "epoch": 0.03, "grad_norm": 1.188730001449585, "learning_rate": 9.9714248673854e-05, "loss": 2.4709, "step": 4410 }, { "epoch": 0.03, "grad_norm": 1.6798044443130493, "learning_rate": 9.971358592723361e-05, "loss": 2.472, "step": 4415 }, { "epoch": 0.03, "grad_norm": 1.4117074012756348, "learning_rate": 9.97129224151525e-05, "loss": 2.423, "step": 4420 }, { "epoch": 0.03, "grad_norm": 56.2673454284668, "learning_rate": 9.971225813762094e-05, "loss": 2.4656, "step": 4425 }, { "epoch": 0.03, "grad_norm": 1.2082257270812988, "learning_rate": 9.971159309464906e-05, "loss": 2.4829, "step": 4430 }, { "epoch": 0.04, "grad_norm": 5.003783702850342, "learning_rate": 9.971092728624718e-05, "loss": 2.4594, "step": 4435 }, { "epoch": 0.04, "grad_norm": 1.2291468381881714, "learning_rate": 9.97102607124255e-05, "loss": 2.4546, "step": 4440 }, { "epoch": 0.04, "grad_norm": 0.5952057838439941, "learning_rate": 9.970959337319433e-05, "loss": 2.4657, "step": 4445 }, { "epoch": 0.04, "grad_norm": 7.087344169616699, "learning_rate": 9.970892526856392e-05, "loss": 2.4644, "step": 4450 }, { "epoch": 0.04, "grad_norm": 1.4197371006011963, "learning_rate": 9.970825639854454e-05, "loss": 2.5678, "step": 4455 }, { "epoch": 0.04, "grad_norm": 5.621391773223877, "learning_rate": 9.970758676314652e-05, "loss": 2.3887, "step": 4460 }, { "epoch": 0.04, "grad_norm": 0.8304689526557922, "learning_rate": 9.970691636238017e-05, "loss": 2.4996, "step": 4465 }, { "epoch": 0.04, "grad_norm": 1.5258510112762451, "learning_rate": 9.970624519625578e-05, "loss": 2.4535, "step": 4470 }, { "epoch": 0.04, "grad_norm": 0.9350457191467285, "learning_rate": 9.970557326478371e-05, "loss": 2.5157, "step": 4475 }, { "epoch": 0.04, "grad_norm": 2.72408127784729, "learning_rate": 9.97049005679743e-05, "loss": 2.5335, "step": 4480 }, { "epoch": 0.04, "grad_norm": 0.5874288082122803, "learning_rate": 9.970422710583791e-05, "loss": 2.4693, "step": 4485 }, { "epoch": 0.04, "grad_norm": 0.9633789658546448, "learning_rate": 9.970355287838491e-05, "loss": 2.4173, "step": 4490 }, { "epoch": 0.04, "grad_norm": 5.302885055541992, "learning_rate": 9.970287788562567e-05, "loss": 2.4612, "step": 4495 }, { "epoch": 0.04, "grad_norm": 1.2865092754364014, "learning_rate": 9.97022021275706e-05, "loss": 2.5685, "step": 4500 }, { "epoch": 0.04, "grad_norm": 4.8376874923706055, "learning_rate": 9.97015256042301e-05, "loss": 2.4736, "step": 4505 }, { "epoch": 0.04, "grad_norm": 0.9332994222640991, "learning_rate": 9.97008483156146e-05, "loss": 2.514, "step": 4510 }, { "epoch": 0.04, "grad_norm": 1.9254236221313477, "learning_rate": 9.970017026173448e-05, "loss": 2.4913, "step": 4515 }, { "epoch": 0.04, "grad_norm": 7.618185520172119, "learning_rate": 9.969949144260022e-05, "loss": 2.4947, "step": 4520 }, { "epoch": 0.04, "grad_norm": 1.0545151233673096, "learning_rate": 9.969881185822227e-05, "loss": 2.5173, "step": 4525 }, { "epoch": 0.04, "grad_norm": 0.6671162843704224, "learning_rate": 9.969813150861111e-05, "loss": 2.5064, "step": 4530 }, { "epoch": 0.04, "grad_norm": 0.7074204683303833, "learning_rate": 9.969745039377717e-05, "loss": 2.514, "step": 4535 }, { "epoch": 0.04, "grad_norm": 3.2114500999450684, "learning_rate": 9.969676851373098e-05, "loss": 2.5814, "step": 4540 }, { "epoch": 0.04, "grad_norm": 1.0560355186462402, "learning_rate": 9.969608586848301e-05, "loss": 2.5097, "step": 4545 }, { "epoch": 0.04, "grad_norm": 0.5754773616790771, "learning_rate": 9.969540245804377e-05, "loss": 2.4462, "step": 4550 }, { "epoch": 0.04, "grad_norm": 0.6706427335739136, "learning_rate": 9.96947182824238e-05, "loss": 2.5842, "step": 4555 }, { "epoch": 0.04, "grad_norm": 0.818139374256134, "learning_rate": 9.969403334163364e-05, "loss": 2.535, "step": 4560 }, { "epoch": 0.04, "grad_norm": 1.1772286891937256, "learning_rate": 9.969334763568383e-05, "loss": 2.5469, "step": 4565 }, { "epoch": 0.04, "grad_norm": 0.8835670948028564, "learning_rate": 9.969266116458491e-05, "loss": 2.5427, "step": 4570 }, { "epoch": 0.04, "grad_norm": 0.9960147142410278, "learning_rate": 9.969197392834747e-05, "loss": 2.5424, "step": 4575 }, { "epoch": 0.04, "grad_norm": 0.8085960149765015, "learning_rate": 9.969128592698207e-05, "loss": 2.5634, "step": 4580 }, { "epoch": 0.04, "grad_norm": 0.6777289509773254, "learning_rate": 9.969059716049934e-05, "loss": 2.419, "step": 4585 }, { "epoch": 0.04, "grad_norm": 0.855076014995575, "learning_rate": 9.968990762890986e-05, "loss": 2.4644, "step": 4590 }, { "epoch": 0.04, "grad_norm": 1.6585725545883179, "learning_rate": 9.968921733222425e-05, "loss": 2.4256, "step": 4595 }, { "epoch": 0.04, "grad_norm": 1.3942941427230835, "learning_rate": 9.968852627045313e-05, "loss": 2.4879, "step": 4600 }, { "epoch": 0.04, "grad_norm": 0.760040283203125, "learning_rate": 9.968783444360716e-05, "loss": 2.5493, "step": 4605 }, { "epoch": 0.04, "grad_norm": 1.7026785612106323, "learning_rate": 9.968714185169698e-05, "loss": 2.5778, "step": 4610 }, { "epoch": 0.04, "grad_norm": 1.067081332206726, "learning_rate": 9.968644849473326e-05, "loss": 2.5441, "step": 4615 }, { "epoch": 0.04, "grad_norm": 15.749674797058105, "learning_rate": 9.968575437272667e-05, "loss": 2.5518, "step": 4620 }, { "epoch": 0.04, "grad_norm": 0.6095540523529053, "learning_rate": 9.968505948568792e-05, "loss": 2.5053, "step": 4625 }, { "epoch": 0.04, "grad_norm": 4.169956684112549, "learning_rate": 9.968436383362766e-05, "loss": 2.4532, "step": 4630 }, { "epoch": 0.04, "grad_norm": 0.7167848348617554, "learning_rate": 9.968366741655663e-05, "loss": 2.4363, "step": 4635 }, { "epoch": 0.04, "grad_norm": 2.0984060764312744, "learning_rate": 9.968297023448558e-05, "loss": 2.3216, "step": 4640 }, { "epoch": 0.04, "grad_norm": 0.9936848282814026, "learning_rate": 9.96822722874252e-05, "loss": 2.4511, "step": 4645 }, { "epoch": 0.04, "grad_norm": 8.994372367858887, "learning_rate": 9.968157357538625e-05, "loss": 2.5109, "step": 4650 }, { "epoch": 0.04, "grad_norm": 0.6544365882873535, "learning_rate": 9.968087409837951e-05, "loss": 2.3831, "step": 4655 }, { "epoch": 0.04, "grad_norm": 1.2369416952133179, "learning_rate": 9.968017385641574e-05, "loss": 2.5168, "step": 4660 }, { "epoch": 0.04, "grad_norm": 1.2204056978225708, "learning_rate": 9.967947284950569e-05, "loss": 2.4545, "step": 4665 }, { "epoch": 0.04, "grad_norm": 1.5159409046173096, "learning_rate": 9.967877107766021e-05, "loss": 2.5937, "step": 4670 }, { "epoch": 0.04, "grad_norm": 9.555320739746094, "learning_rate": 9.967806854089006e-05, "loss": 2.534, "step": 4675 }, { "epoch": 0.04, "grad_norm": 0.6831074953079224, "learning_rate": 9.967736523920607e-05, "loss": 2.5057, "step": 4680 }, { "epoch": 0.04, "grad_norm": 4.3828206062316895, "learning_rate": 9.967666117261907e-05, "loss": 2.5505, "step": 4685 }, { "epoch": 0.04, "grad_norm": 0.9608336091041565, "learning_rate": 9.96759563411399e-05, "loss": 2.5324, "step": 4690 }, { "epoch": 0.04, "grad_norm": 1.587972640991211, "learning_rate": 9.967525074477943e-05, "loss": 2.492, "step": 4695 }, { "epoch": 0.04, "grad_norm": 0.7100768089294434, "learning_rate": 9.96745443835485e-05, "loss": 2.4522, "step": 4700 }, { "epoch": 0.04, "grad_norm": 0.6527875065803528, "learning_rate": 9.967383725745799e-05, "loss": 2.4604, "step": 4705 }, { "epoch": 0.04, "grad_norm": 0.8383848071098328, "learning_rate": 9.96731293665188e-05, "loss": 2.3556, "step": 4710 }, { "epoch": 0.04, "grad_norm": 0.6354039907455444, "learning_rate": 9.967242071074181e-05, "loss": 2.5264, "step": 4715 }, { "epoch": 0.04, "grad_norm": 0.9621263742446899, "learning_rate": 9.967171129013797e-05, "loss": 2.5203, "step": 4720 }, { "epoch": 0.04, "grad_norm": 1.3833376169204712, "learning_rate": 9.967100110471817e-05, "loss": 2.4875, "step": 4725 }, { "epoch": 0.04, "grad_norm": 1.9356681108474731, "learning_rate": 9.967029015449334e-05, "loss": 2.4928, "step": 4730 }, { "epoch": 0.04, "grad_norm": 0.7303605675697327, "learning_rate": 9.966957843947444e-05, "loss": 2.4957, "step": 4735 }, { "epoch": 0.04, "grad_norm": 0.9117994904518127, "learning_rate": 9.966886595967243e-05, "loss": 2.4434, "step": 4740 }, { "epoch": 0.04, "grad_norm": 9.138729095458984, "learning_rate": 9.966815271509828e-05, "loss": 2.4774, "step": 4745 }, { "epoch": 0.04, "grad_norm": 0.8541014194488525, "learning_rate": 9.966743870576297e-05, "loss": 2.5039, "step": 4750 }, { "epoch": 0.04, "grad_norm": 2.4821598529815674, "learning_rate": 9.96667239316775e-05, "loss": 2.5114, "step": 4755 }, { "epoch": 0.04, "grad_norm": 11.643621444702148, "learning_rate": 9.966600839285286e-05, "loss": 2.5208, "step": 4760 }, { "epoch": 0.04, "grad_norm": 1.0150034427642822, "learning_rate": 9.966529208930008e-05, "loss": 2.3943, "step": 4765 }, { "epoch": 0.04, "grad_norm": 0.5926784873008728, "learning_rate": 9.96645750210302e-05, "loss": 2.4419, "step": 4770 }, { "epoch": 0.04, "grad_norm": 1.2746299505233765, "learning_rate": 9.966385718805423e-05, "loss": 2.4771, "step": 4775 }, { "epoch": 0.04, "grad_norm": 1.4886510372161865, "learning_rate": 9.966313859038325e-05, "loss": 2.3957, "step": 4780 }, { "epoch": 0.04, "grad_norm": 0.7025569677352905, "learning_rate": 9.966241922802832e-05, "loss": 2.5153, "step": 4785 }, { "epoch": 0.04, "grad_norm": 0.6843971610069275, "learning_rate": 9.96616991010005e-05, "loss": 2.4538, "step": 4790 }, { "epoch": 0.04, "grad_norm": 9.501198768615723, "learning_rate": 9.96609782093109e-05, "loss": 2.4011, "step": 4795 }, { "epoch": 0.04, "grad_norm": 0.6182028651237488, "learning_rate": 9.96602565529706e-05, "loss": 2.4529, "step": 4800 }, { "epoch": 0.04, "grad_norm": 1.7064813375473022, "learning_rate": 9.965953413199071e-05, "loss": 2.4614, "step": 4805 }, { "epoch": 0.04, "grad_norm": 0.6552773714065552, "learning_rate": 9.965881094638238e-05, "loss": 2.5696, "step": 4810 }, { "epoch": 0.04, "grad_norm": 6.229384422302246, "learning_rate": 9.965808699615674e-05, "loss": 2.5547, "step": 4815 }, { "epoch": 0.04, "grad_norm": 0.9367731213569641, "learning_rate": 9.965736228132492e-05, "loss": 2.4406, "step": 4820 }, { "epoch": 0.04, "grad_norm": 0.9215686917304993, "learning_rate": 9.965663680189807e-05, "loss": 2.5307, "step": 4825 }, { "epoch": 0.04, "grad_norm": 2.022834539413452, "learning_rate": 9.965591055788738e-05, "loss": 2.4893, "step": 4830 }, { "epoch": 0.04, "grad_norm": 0.6709932088851929, "learning_rate": 9.965518354930405e-05, "loss": 2.4526, "step": 4835 }, { "epoch": 0.04, "grad_norm": 1.158338189125061, "learning_rate": 9.965445577615922e-05, "loss": 2.5482, "step": 4840 }, { "epoch": 0.04, "grad_norm": 1.2223151922225952, "learning_rate": 9.965372723846415e-05, "loss": 2.4383, "step": 4845 }, { "epoch": 0.04, "grad_norm": 1.42411208152771, "learning_rate": 9.965299793623003e-05, "loss": 2.5684, "step": 4850 }, { "epoch": 0.04, "grad_norm": 0.7620886564254761, "learning_rate": 9.96522678694681e-05, "loss": 2.5228, "step": 4855 }, { "epoch": 0.04, "grad_norm": 1.2974141836166382, "learning_rate": 9.965153703818959e-05, "loss": 2.4935, "step": 4860 }, { "epoch": 0.04, "grad_norm": 1.1414613723754883, "learning_rate": 9.965080544240575e-05, "loss": 2.5033, "step": 4865 }, { "epoch": 0.04, "grad_norm": 0.8289280533790588, "learning_rate": 9.965007308212786e-05, "loss": 2.3917, "step": 4870 }, { "epoch": 0.04, "grad_norm": 0.696593165397644, "learning_rate": 9.964933995736721e-05, "loss": 2.5358, "step": 4875 }, { "epoch": 0.04, "grad_norm": 0.7842949628829956, "learning_rate": 9.964860606813504e-05, "loss": 2.373, "step": 4880 }, { "epoch": 0.04, "grad_norm": 1.269083023071289, "learning_rate": 9.96478714144427e-05, "loss": 2.5144, "step": 4885 }, { "epoch": 0.04, "grad_norm": 0.8201007843017578, "learning_rate": 9.964713599630146e-05, "loss": 2.3411, "step": 4890 }, { "epoch": 0.04, "grad_norm": 1.0418899059295654, "learning_rate": 9.964639981372267e-05, "loss": 2.4628, "step": 4895 }, { "epoch": 0.04, "grad_norm": 0.6437423825263977, "learning_rate": 9.964566286671766e-05, "loss": 2.4535, "step": 4900 }, { "epoch": 0.04, "grad_norm": 6.69949197769165, "learning_rate": 9.964492515529777e-05, "loss": 2.4544, "step": 4905 }, { "epoch": 0.04, "grad_norm": 0.5273777842521667, "learning_rate": 9.964418667947438e-05, "loss": 2.4234, "step": 4910 }, { "epoch": 0.04, "grad_norm": 1.136296033859253, "learning_rate": 9.964344743925884e-05, "loss": 2.3746, "step": 4915 }, { "epoch": 0.04, "grad_norm": 0.753968358039856, "learning_rate": 9.964270743466253e-05, "loss": 2.4749, "step": 4920 }, { "epoch": 0.04, "grad_norm": 1.619755744934082, "learning_rate": 9.964196666569686e-05, "loss": 2.5206, "step": 4925 }, { "epoch": 0.04, "grad_norm": 1.7213090658187866, "learning_rate": 9.964122513237322e-05, "loss": 2.4916, "step": 4930 }, { "epoch": 0.04, "grad_norm": 0.8247066736221313, "learning_rate": 9.964048283470305e-05, "loss": 2.3709, "step": 4935 }, { "epoch": 0.04, "grad_norm": 0.7608567476272583, "learning_rate": 9.963973977269776e-05, "loss": 2.466, "step": 4940 }, { "epoch": 0.04, "grad_norm": 0.5843170881271362, "learning_rate": 9.963899594636879e-05, "loss": 2.4745, "step": 4945 }, { "epoch": 0.04, "grad_norm": 0.6609501838684082, "learning_rate": 9.96382513557276e-05, "loss": 2.3252, "step": 4950 }, { "epoch": 0.04, "grad_norm": 0.5760492086410522, "learning_rate": 9.963750600078565e-05, "loss": 2.44, "step": 4955 }, { "epoch": 0.04, "grad_norm": 0.726439893245697, "learning_rate": 9.963675988155443e-05, "loss": 2.5057, "step": 4960 }, { "epoch": 0.04, "grad_norm": 0.7160363793373108, "learning_rate": 9.963601299804541e-05, "loss": 2.3867, "step": 4965 }, { "epoch": 0.04, "grad_norm": 1.1992379426956177, "learning_rate": 9.963526535027012e-05, "loss": 2.4054, "step": 4970 }, { "epoch": 0.04, "grad_norm": 0.8473523855209351, "learning_rate": 9.963451693824003e-05, "loss": 2.5606, "step": 4975 }, { "epoch": 0.04, "grad_norm": 1.0527797937393188, "learning_rate": 9.963376776196669e-05, "loss": 2.4721, "step": 4980 }, { "epoch": 0.04, "grad_norm": 1.899242877960205, "learning_rate": 9.963301782146163e-05, "loss": 2.4227, "step": 4985 }, { "epoch": 0.04, "grad_norm": 0.7237558364868164, "learning_rate": 9.96322671167364e-05, "loss": 2.4451, "step": 4990 }, { "epoch": 0.04, "grad_norm": 1.3388500213623047, "learning_rate": 9.963151564780256e-05, "loss": 2.4373, "step": 4995 }, { "epoch": 0.04, "grad_norm": 1.5860034227371216, "learning_rate": 9.963076341467165e-05, "loss": 2.3375, "step": 5000 }, { "epoch": 0.04, "grad_norm": 0.8965621590614319, "learning_rate": 9.963001041735531e-05, "loss": 2.4936, "step": 5005 }, { "epoch": 0.04, "grad_norm": 1.1495734453201294, "learning_rate": 9.962925665586509e-05, "loss": 2.4504, "step": 5010 }, { "epoch": 0.04, "grad_norm": 0.7743747234344482, "learning_rate": 9.962850213021262e-05, "loss": 2.4506, "step": 5015 }, { "epoch": 0.04, "grad_norm": 1.8350714445114136, "learning_rate": 9.96277468404095e-05, "loss": 2.4993, "step": 5020 }, { "epoch": 0.04, "grad_norm": 0.7750059366226196, "learning_rate": 9.962699078646735e-05, "loss": 2.5333, "step": 5025 }, { "epoch": 0.04, "grad_norm": 0.6362027525901794, "learning_rate": 9.962623396839785e-05, "loss": 2.423, "step": 5030 }, { "epoch": 0.04, "grad_norm": 1.9829604625701904, "learning_rate": 9.962547638621262e-05, "loss": 2.54, "step": 5035 }, { "epoch": 0.04, "grad_norm": 0.8062134981155396, "learning_rate": 9.962471803992334e-05, "loss": 2.4895, "step": 5040 }, { "epoch": 0.04, "grad_norm": 0.695792555809021, "learning_rate": 9.962395892954168e-05, "loss": 2.3401, "step": 5045 }, { "epoch": 0.04, "grad_norm": 1.4658273458480835, "learning_rate": 9.962319905507934e-05, "loss": 2.4398, "step": 5050 }, { "epoch": 0.04, "grad_norm": 0.6275031566619873, "learning_rate": 9.9622438416548e-05, "loss": 2.4664, "step": 5055 }, { "epoch": 0.04, "grad_norm": 4.324435710906982, "learning_rate": 9.962167701395937e-05, "loss": 2.4079, "step": 5060 }, { "epoch": 0.04, "grad_norm": 1.4788026809692383, "learning_rate": 9.962091484732521e-05, "loss": 2.5545, "step": 5065 }, { "epoch": 0.04, "grad_norm": 0.8761649131774902, "learning_rate": 9.962015191665722e-05, "loss": 2.4059, "step": 5070 }, { "epoch": 0.04, "grad_norm": 4.358525276184082, "learning_rate": 9.961938822196716e-05, "loss": 2.5032, "step": 5075 }, { "epoch": 0.04, "grad_norm": 2.2021636962890625, "learning_rate": 9.961862376326678e-05, "loss": 2.4338, "step": 5080 }, { "epoch": 0.04, "grad_norm": 0.6921555995941162, "learning_rate": 9.961785854056788e-05, "loss": 2.4369, "step": 5085 }, { "epoch": 0.04, "grad_norm": 0.5950436592102051, "learning_rate": 9.961709255388221e-05, "loss": 2.3319, "step": 5090 }, { "epoch": 0.04, "grad_norm": 1.3542665243148804, "learning_rate": 9.961632580322156e-05, "loss": 2.4158, "step": 5095 }, { "epoch": 0.04, "grad_norm": 1.3912501335144043, "learning_rate": 9.961555828859777e-05, "loss": 2.5051, "step": 5100 }, { "epoch": 0.04, "grad_norm": 1.3188029527664185, "learning_rate": 9.961479001002264e-05, "loss": 2.4457, "step": 5105 }, { "epoch": 0.04, "grad_norm": 1.1079611778259277, "learning_rate": 9.961402096750799e-05, "loss": 2.4547, "step": 5110 }, { "epoch": 0.04, "grad_norm": 0.9491347670555115, "learning_rate": 9.961325116106566e-05, "loss": 2.4543, "step": 5115 }, { "epoch": 0.04, "grad_norm": 0.7553776502609253, "learning_rate": 9.961248059070751e-05, "loss": 2.5005, "step": 5120 }, { "epoch": 0.04, "grad_norm": 0.5640578866004944, "learning_rate": 9.961170925644542e-05, "loss": 2.4035, "step": 5125 }, { "epoch": 0.04, "grad_norm": 1.7542049884796143, "learning_rate": 9.961093715829127e-05, "loss": 2.4841, "step": 5130 }, { "epoch": 0.04, "grad_norm": 0.9023405313491821, "learning_rate": 9.961016429625691e-05, "loss": 2.482, "step": 5135 }, { "epoch": 0.04, "grad_norm": 1.1788444519042969, "learning_rate": 9.960939067035426e-05, "loss": 2.4123, "step": 5140 }, { "epoch": 0.04, "grad_norm": 0.5944883823394775, "learning_rate": 9.960861628059524e-05, "loss": 2.3022, "step": 5145 }, { "epoch": 0.04, "grad_norm": 1.4618780612945557, "learning_rate": 9.960784112699177e-05, "loss": 2.4284, "step": 5150 }, { "epoch": 0.04, "grad_norm": 2.9487245082855225, "learning_rate": 9.960706520955578e-05, "loss": 2.405, "step": 5155 }, { "epoch": 0.04, "grad_norm": 0.9469388723373413, "learning_rate": 9.960628852829923e-05, "loss": 2.5301, "step": 5160 }, { "epoch": 0.04, "grad_norm": 28.036903381347656, "learning_rate": 9.960551108323406e-05, "loss": 2.5219, "step": 5165 }, { "epoch": 0.04, "grad_norm": 0.752023458480835, "learning_rate": 9.960473287437224e-05, "loss": 2.438, "step": 5170 }, { "epoch": 0.04, "grad_norm": 20.4979190826416, "learning_rate": 9.960395390172579e-05, "loss": 2.5024, "step": 5175 }, { "epoch": 0.04, "grad_norm": 1.1260969638824463, "learning_rate": 9.960317416530665e-05, "loss": 2.6393, "step": 5180 }, { "epoch": 0.04, "grad_norm": 0.6956133842468262, "learning_rate": 9.960239366512686e-05, "loss": 2.4678, "step": 5185 }, { "epoch": 0.04, "grad_norm": 0.7920162081718445, "learning_rate": 9.960161240119843e-05, "loss": 2.4245, "step": 5190 }, { "epoch": 0.04, "grad_norm": 1.270247220993042, "learning_rate": 9.96008303735334e-05, "loss": 2.5481, "step": 5195 }, { "epoch": 0.04, "grad_norm": 47.3443603515625, "learning_rate": 9.960004758214379e-05, "loss": 2.4571, "step": 5200 }, { "epoch": 0.04, "grad_norm": 0.9950844645500183, "learning_rate": 9.959926402704165e-05, "loss": 2.5308, "step": 5205 }, { "epoch": 0.04, "grad_norm": 2.5294418334960938, "learning_rate": 9.959847970823908e-05, "loss": 2.4808, "step": 5210 }, { "epoch": 0.04, "grad_norm": 0.9595973491668701, "learning_rate": 9.959769462574813e-05, "loss": 2.3974, "step": 5215 }, { "epoch": 0.04, "grad_norm": 0.8568931221961975, "learning_rate": 9.959690877958088e-05, "loss": 2.4091, "step": 5220 }, { "epoch": 0.04, "grad_norm": 8.183117866516113, "learning_rate": 9.959612216974946e-05, "loss": 2.3808, "step": 5225 }, { "epoch": 0.04, "grad_norm": 1.0979526042938232, "learning_rate": 9.959533479626596e-05, "loss": 2.5401, "step": 5230 }, { "epoch": 0.04, "grad_norm": 0.628634512424469, "learning_rate": 9.95945466591425e-05, "loss": 2.5484, "step": 5235 }, { "epoch": 0.04, "grad_norm": 0.9426279067993164, "learning_rate": 9.959375775839124e-05, "loss": 2.4342, "step": 5240 }, { "epoch": 0.04, "grad_norm": 1.2173657417297363, "learning_rate": 9.95929680940243e-05, "loss": 2.4266, "step": 5245 }, { "epoch": 0.04, "grad_norm": 11.166548728942871, "learning_rate": 9.959217766605384e-05, "loss": 2.4008, "step": 5250 }, { "epoch": 0.04, "grad_norm": 0.5485805869102478, "learning_rate": 9.959138647449205e-05, "loss": 2.4792, "step": 5255 }, { "epoch": 0.04, "grad_norm": 3.2625975608825684, "learning_rate": 9.95905945193511e-05, "loss": 2.4351, "step": 5260 }, { "epoch": 0.04, "grad_norm": 0.8219376802444458, "learning_rate": 9.958980180064318e-05, "loss": 2.4767, "step": 5265 }, { "epoch": 0.04, "grad_norm": 1.5046974420547485, "learning_rate": 9.958900831838051e-05, "loss": 2.4842, "step": 5270 }, { "epoch": 0.04, "grad_norm": 1.2624503374099731, "learning_rate": 9.958821407257531e-05, "loss": 2.4711, "step": 5275 }, { "epoch": 0.04, "grad_norm": 2.195519208908081, "learning_rate": 9.958741906323978e-05, "loss": 2.4613, "step": 5280 }, { "epoch": 0.04, "grad_norm": 21.471158981323242, "learning_rate": 9.95866232903862e-05, "loss": 2.5165, "step": 5285 }, { "epoch": 0.04, "grad_norm": 1.011244773864746, "learning_rate": 9.958582675402677e-05, "loss": 2.4362, "step": 5290 }, { "epoch": 0.04, "grad_norm": 0.751957893371582, "learning_rate": 9.958502945417382e-05, "loss": 2.4691, "step": 5295 }, { "epoch": 0.04, "grad_norm": 0.8029015064239502, "learning_rate": 9.958423139083958e-05, "loss": 2.5458, "step": 5300 }, { "epoch": 0.04, "grad_norm": 2.4250776767730713, "learning_rate": 9.958343256403635e-05, "loss": 2.3733, "step": 5305 }, { "epoch": 0.04, "grad_norm": 2.2336785793304443, "learning_rate": 9.958263297377643e-05, "loss": 2.4981, "step": 5310 }, { "epoch": 0.04, "grad_norm": 0.8627798557281494, "learning_rate": 9.958183262007215e-05, "loss": 2.4682, "step": 5315 }, { "epoch": 0.04, "grad_norm": 0.6489814519882202, "learning_rate": 9.95810315029358e-05, "loss": 2.4504, "step": 5320 }, { "epoch": 0.04, "grad_norm": 0.8141399621963501, "learning_rate": 9.958022962237974e-05, "loss": 2.3238, "step": 5325 }, { "epoch": 0.04, "grad_norm": 3.039771556854248, "learning_rate": 9.957942697841632e-05, "loss": 2.5238, "step": 5330 }, { "epoch": 0.04, "grad_norm": 1.4041558504104614, "learning_rate": 9.957862357105785e-05, "loss": 2.3927, "step": 5335 }, { "epoch": 0.04, "grad_norm": 0.9037558436393738, "learning_rate": 9.957781940031676e-05, "loss": 2.4674, "step": 5340 }, { "epoch": 0.04, "grad_norm": 1.2327488660812378, "learning_rate": 9.957701446620543e-05, "loss": 2.4402, "step": 5345 }, { "epoch": 0.04, "grad_norm": 0.959912896156311, "learning_rate": 9.957620876873621e-05, "loss": 2.513, "step": 5350 }, { "epoch": 0.04, "grad_norm": 0.82025146484375, "learning_rate": 9.957540230792152e-05, "loss": 2.4582, "step": 5355 }, { "epoch": 0.04, "grad_norm": 0.7162131667137146, "learning_rate": 9.957459508377381e-05, "loss": 2.4113, "step": 5360 }, { "epoch": 0.04, "grad_norm": 2.131065845489502, "learning_rate": 9.957378709630548e-05, "loss": 2.5119, "step": 5365 }, { "epoch": 0.04, "grad_norm": 1.1246188879013062, "learning_rate": 9.957297834552898e-05, "loss": 2.4952, "step": 5370 }, { "epoch": 0.04, "grad_norm": 24.261186599731445, "learning_rate": 9.957216883145675e-05, "loss": 2.3655, "step": 5375 }, { "epoch": 0.04, "grad_norm": 0.6341184973716736, "learning_rate": 9.957135855410127e-05, "loss": 2.3407, "step": 5380 }, { "epoch": 0.04, "grad_norm": 0.8851598501205444, "learning_rate": 9.957054751347501e-05, "loss": 2.4988, "step": 5385 }, { "epoch": 0.04, "grad_norm": 5.903181552886963, "learning_rate": 9.956973570959045e-05, "loss": 2.5152, "step": 5390 }, { "epoch": 0.04, "grad_norm": 0.9890641570091248, "learning_rate": 9.956892314246012e-05, "loss": 2.4276, "step": 5395 }, { "epoch": 0.04, "grad_norm": 12.017911911010742, "learning_rate": 9.956810981209647e-05, "loss": 2.4909, "step": 5400 }, { "epoch": 0.04, "grad_norm": 0.8295998573303223, "learning_rate": 9.956729571851211e-05, "loss": 2.4837, "step": 5405 }, { "epoch": 0.04, "grad_norm": 3.9519195556640625, "learning_rate": 9.956648086171949e-05, "loss": 2.4286, "step": 5410 }, { "epoch": 0.04, "grad_norm": 4.8375749588012695, "learning_rate": 9.956566524173121e-05, "loss": 2.4999, "step": 5415 }, { "epoch": 0.04, "grad_norm": 1.29248046875, "learning_rate": 9.956484885855982e-05, "loss": 2.5473, "step": 5420 }, { "epoch": 0.04, "grad_norm": 0.713558554649353, "learning_rate": 9.956403171221786e-05, "loss": 2.3735, "step": 5425 }, { "epoch": 0.04, "grad_norm": 0.6501023173332214, "learning_rate": 9.956321380271796e-05, "loss": 2.4406, "step": 5430 }, { "epoch": 0.04, "grad_norm": 0.6077499985694885, "learning_rate": 9.956239513007268e-05, "loss": 2.4159, "step": 5435 }, { "epoch": 0.04, "grad_norm": 0.6564694046974182, "learning_rate": 9.956157569429464e-05, "loss": 2.4999, "step": 5440 }, { "epoch": 0.04, "grad_norm": 1.8838554620742798, "learning_rate": 9.956075549539643e-05, "loss": 2.4414, "step": 5445 }, { "epoch": 0.04, "grad_norm": 1.316749930381775, "learning_rate": 9.955993453339072e-05, "loss": 2.478, "step": 5450 }, { "epoch": 0.04, "grad_norm": 0.8767583966255188, "learning_rate": 9.955911280829011e-05, "loss": 2.3964, "step": 5455 }, { "epoch": 0.04, "grad_norm": 4.961359024047852, "learning_rate": 9.95582903201073e-05, "loss": 2.4365, "step": 5460 }, { "epoch": 0.04, "grad_norm": 0.8863642811775208, "learning_rate": 9.955746706885492e-05, "loss": 2.4374, "step": 5465 }, { "epoch": 0.04, "grad_norm": 0.7299385666847229, "learning_rate": 9.955664305454564e-05, "loss": 2.4141, "step": 5470 }, { "epoch": 0.04, "grad_norm": 1.4970512390136719, "learning_rate": 9.955581827719216e-05, "loss": 2.4493, "step": 5475 }, { "epoch": 0.04, "grad_norm": 0.8494284749031067, "learning_rate": 9.955499273680719e-05, "loss": 2.4104, "step": 5480 }, { "epoch": 0.04, "grad_norm": 0.9648477435112, "learning_rate": 9.955416643340343e-05, "loss": 2.4418, "step": 5485 }, { "epoch": 0.04, "grad_norm": 0.6430608034133911, "learning_rate": 9.95533393669936e-05, "loss": 2.4687, "step": 5490 }, { "epoch": 0.04, "grad_norm": 0.8171415328979492, "learning_rate": 9.955251153759044e-05, "loss": 2.445, "step": 5495 }, { "epoch": 0.04, "grad_norm": 2.7282419204711914, "learning_rate": 9.955168294520672e-05, "loss": 2.527, "step": 5500 }, { "epoch": 0.04, "grad_norm": 1.3247615098953247, "learning_rate": 9.955085358985514e-05, "loss": 2.3285, "step": 5505 }, { "epoch": 0.04, "grad_norm": 0.6209290623664856, "learning_rate": 9.955002347154851e-05, "loss": 2.4624, "step": 5510 }, { "epoch": 0.04, "grad_norm": 5.925119400024414, "learning_rate": 9.954919259029962e-05, "loss": 2.3937, "step": 5515 }, { "epoch": 0.04, "grad_norm": 1.1366395950317383, "learning_rate": 9.954836094612123e-05, "loss": 2.4668, "step": 5520 }, { "epoch": 0.04, "grad_norm": 2.1967954635620117, "learning_rate": 9.954752853902619e-05, "loss": 2.3883, "step": 5525 }, { "epoch": 0.04, "grad_norm": 0.5749600529670715, "learning_rate": 9.954669536902728e-05, "loss": 2.3545, "step": 5530 }, { "epoch": 0.04, "grad_norm": 1.8527699708938599, "learning_rate": 9.954586143613733e-05, "loss": 2.4397, "step": 5535 }, { "epoch": 0.04, "grad_norm": 3.255234479904175, "learning_rate": 9.954502674036919e-05, "loss": 2.467, "step": 5540 }, { "epoch": 0.04, "grad_norm": 0.9579917192459106, "learning_rate": 9.954419128173572e-05, "loss": 2.4469, "step": 5545 }, { "epoch": 0.04, "grad_norm": 0.8501748442649841, "learning_rate": 9.954335506024978e-05, "loss": 2.5259, "step": 5550 }, { "epoch": 0.04, "grad_norm": 0.7592048048973083, "learning_rate": 9.954251807592423e-05, "loss": 2.4499, "step": 5555 }, { "epoch": 0.04, "grad_norm": 0.8333644866943359, "learning_rate": 9.954168032877198e-05, "loss": 2.4944, "step": 5560 }, { "epoch": 0.04, "grad_norm": 0.775357186794281, "learning_rate": 9.95408418188059e-05, "loss": 2.4623, "step": 5565 }, { "epoch": 0.04, "grad_norm": 1.6152263879776, "learning_rate": 9.954000254603894e-05, "loss": 2.4934, "step": 5570 }, { "epoch": 0.04, "grad_norm": 0.6403139233589172, "learning_rate": 9.9539162510484e-05, "loss": 2.4549, "step": 5575 }, { "epoch": 0.04, "grad_norm": 2.330981969833374, "learning_rate": 9.953832171215401e-05, "loss": 2.4125, "step": 5580 }, { "epoch": 0.04, "grad_norm": 1.2553397417068481, "learning_rate": 9.953748015106193e-05, "loss": 2.3735, "step": 5585 }, { "epoch": 0.04, "grad_norm": 0.6711847186088562, "learning_rate": 9.953663782722069e-05, "loss": 2.3783, "step": 5590 }, { "epoch": 0.04, "grad_norm": 1.4691882133483887, "learning_rate": 9.953579474064329e-05, "loss": 2.3701, "step": 5595 }, { "epoch": 0.04, "grad_norm": 0.8467762470245361, "learning_rate": 9.95349508913427e-05, "loss": 2.4742, "step": 5600 }, { "epoch": 0.04, "grad_norm": 1.5817852020263672, "learning_rate": 9.953410627933193e-05, "loss": 2.4051, "step": 5605 }, { "epoch": 0.04, "grad_norm": 0.7455315589904785, "learning_rate": 9.953326090462397e-05, "loss": 2.4727, "step": 5610 }, { "epoch": 0.04, "grad_norm": 2.0927181243896484, "learning_rate": 9.953241476723181e-05, "loss": 2.4724, "step": 5615 }, { "epoch": 0.04, "grad_norm": 0.7695971131324768, "learning_rate": 9.953156786716851e-05, "loss": 2.4228, "step": 5620 }, { "epoch": 0.04, "grad_norm": 8.26998233795166, "learning_rate": 9.953072020444711e-05, "loss": 2.4539, "step": 5625 }, { "epoch": 0.04, "grad_norm": 0.6999443173408508, "learning_rate": 9.952987177908066e-05, "loss": 2.4655, "step": 5630 }, { "epoch": 0.04, "grad_norm": 4.575010776519775, "learning_rate": 9.95290225910822e-05, "loss": 2.4165, "step": 5635 }, { "epoch": 0.04, "grad_norm": 0.6047275066375732, "learning_rate": 9.952817264046486e-05, "loss": 2.4173, "step": 5640 }, { "epoch": 0.04, "grad_norm": 0.6233522891998291, "learning_rate": 9.952732192724167e-05, "loss": 2.4024, "step": 5645 }, { "epoch": 0.04, "grad_norm": 1.1003082990646362, "learning_rate": 9.952647045142576e-05, "loss": 2.4863, "step": 5650 }, { "epoch": 0.04, "grad_norm": 0.7332708239555359, "learning_rate": 9.952561821303021e-05, "loss": 2.4601, "step": 5655 }, { "epoch": 0.04, "grad_norm": 1.1781197786331177, "learning_rate": 9.952476521206818e-05, "loss": 2.4485, "step": 5660 }, { "epoch": 0.04, "grad_norm": 2.4591927528381348, "learning_rate": 9.952391144855278e-05, "loss": 2.3531, "step": 5665 }, { "epoch": 0.04, "grad_norm": 0.6761226058006287, "learning_rate": 9.952305692249717e-05, "loss": 2.4228, "step": 5670 }, { "epoch": 0.04, "grad_norm": 0.8278024196624756, "learning_rate": 9.952220163391451e-05, "loss": 2.3658, "step": 5675 }, { "epoch": 0.04, "grad_norm": 1.6435962915420532, "learning_rate": 9.952134558281793e-05, "loss": 2.3473, "step": 5680 }, { "epoch": 0.04, "grad_norm": 1.3481897115707397, "learning_rate": 9.952048876922066e-05, "loss": 2.4307, "step": 5685 }, { "epoch": 0.04, "grad_norm": 1.012253761291504, "learning_rate": 9.951963119313588e-05, "loss": 2.2894, "step": 5690 }, { "epoch": 0.04, "grad_norm": 1.6459078788757324, "learning_rate": 9.951877285457678e-05, "loss": 2.4483, "step": 5695 }, { "epoch": 0.04, "grad_norm": 2.9752230644226074, "learning_rate": 9.951791375355659e-05, "loss": 2.5181, "step": 5700 }, { "epoch": 0.05, "grad_norm": 1.2505334615707397, "learning_rate": 9.951705389008852e-05, "loss": 2.3599, "step": 5705 }, { "epoch": 0.05, "grad_norm": 0.6725990772247314, "learning_rate": 9.951619326418582e-05, "loss": 2.5307, "step": 5710 }, { "epoch": 0.05, "grad_norm": 0.785463809967041, "learning_rate": 9.951533187586178e-05, "loss": 2.3976, "step": 5715 }, { "epoch": 0.05, "grad_norm": 1.1254979372024536, "learning_rate": 9.951446972512958e-05, "loss": 2.4919, "step": 5720 }, { "epoch": 0.05, "grad_norm": 10.543249130249023, "learning_rate": 9.951360681200256e-05, "loss": 2.4458, "step": 5725 }, { "epoch": 0.05, "grad_norm": 1.3127411603927612, "learning_rate": 9.9512743136494e-05, "loss": 2.4274, "step": 5730 }, { "epoch": 0.05, "grad_norm": 3.076998472213745, "learning_rate": 9.951187869861715e-05, "loss": 2.4256, "step": 5735 }, { "epoch": 0.05, "grad_norm": 1.0276529788970947, "learning_rate": 9.951101349838538e-05, "loss": 2.3587, "step": 5740 }, { "epoch": 0.05, "grad_norm": 1.2900047302246094, "learning_rate": 9.9510147535812e-05, "loss": 2.3316, "step": 5745 }, { "epoch": 0.05, "grad_norm": 0.8634889721870422, "learning_rate": 9.950928081091032e-05, "loss": 2.4297, "step": 5750 }, { "epoch": 0.05, "grad_norm": 0.6170324087142944, "learning_rate": 9.95084133236937e-05, "loss": 2.464, "step": 5755 }, { "epoch": 0.05, "grad_norm": 1.0226573944091797, "learning_rate": 9.950754507417547e-05, "loss": 2.4964, "step": 5760 }, { "epoch": 0.05, "grad_norm": 0.6896876096725464, "learning_rate": 9.950667606236906e-05, "loss": 2.4747, "step": 5765 }, { "epoch": 0.05, "grad_norm": 0.8178924918174744, "learning_rate": 9.95058062882878e-05, "loss": 2.5239, "step": 5770 }, { "epoch": 0.05, "grad_norm": 6.6760053634643555, "learning_rate": 9.950493575194509e-05, "loss": 2.4229, "step": 5775 }, { "epoch": 0.05, "grad_norm": 0.67583167552948, "learning_rate": 9.950406445335434e-05, "loss": 2.4463, "step": 5780 }, { "epoch": 0.05, "grad_norm": 0.827716052532196, "learning_rate": 9.950319239252896e-05, "loss": 2.4543, "step": 5785 }, { "epoch": 0.05, "grad_norm": 0.625660240650177, "learning_rate": 9.95023195694824e-05, "loss": 2.3517, "step": 5790 }, { "epoch": 0.05, "grad_norm": 2.4658234119415283, "learning_rate": 9.950144598422809e-05, "loss": 2.4789, "step": 5795 }, { "epoch": 0.05, "grad_norm": 0.918131947517395, "learning_rate": 9.950057163677944e-05, "loss": 2.4182, "step": 5800 }, { "epoch": 0.05, "grad_norm": 0.9638584852218628, "learning_rate": 9.949969652714997e-05, "loss": 2.409, "step": 5805 }, { "epoch": 0.05, "grad_norm": 1.8958684206008911, "learning_rate": 9.949882065535312e-05, "loss": 2.4427, "step": 5810 }, { "epoch": 0.05, "grad_norm": 4.556503772735596, "learning_rate": 9.949794402140239e-05, "loss": 2.489, "step": 5815 }, { "epoch": 0.05, "grad_norm": 0.5943834781646729, "learning_rate": 9.949706662531127e-05, "loss": 2.4098, "step": 5820 }, { "epoch": 0.05, "grad_norm": 1.0498472452163696, "learning_rate": 9.949618846709327e-05, "loss": 2.4834, "step": 5825 }, { "epoch": 0.05, "grad_norm": 5.513410568237305, "learning_rate": 9.949530954676194e-05, "loss": 2.4144, "step": 5830 }, { "epoch": 0.05, "grad_norm": 0.6448653340339661, "learning_rate": 9.949442986433076e-05, "loss": 2.3569, "step": 5835 }, { "epoch": 0.05, "grad_norm": 1.30804443359375, "learning_rate": 9.949354941981332e-05, "loss": 2.4337, "step": 5840 }, { "epoch": 0.05, "grad_norm": 2.7494187355041504, "learning_rate": 9.949266821322315e-05, "loss": 2.4231, "step": 5845 }, { "epoch": 0.05, "grad_norm": 1.1154729127883911, "learning_rate": 9.949178624457384e-05, "loss": 2.439, "step": 5850 }, { "epoch": 0.05, "grad_norm": 2.7702255249023438, "learning_rate": 9.949090351387895e-05, "loss": 2.3689, "step": 5855 }, { "epoch": 0.05, "grad_norm": 1.0526078939437866, "learning_rate": 9.949002002115208e-05, "loss": 2.4971, "step": 5860 }, { "epoch": 0.05, "grad_norm": 0.7811078429222107, "learning_rate": 9.948913576640683e-05, "loss": 2.4465, "step": 5865 }, { "epoch": 0.05, "grad_norm": 2.4182822704315186, "learning_rate": 9.948825074965682e-05, "loss": 2.4001, "step": 5870 }, { "epoch": 0.05, "grad_norm": 1.0558136701583862, "learning_rate": 9.948736497091569e-05, "loss": 2.434, "step": 5875 }, { "epoch": 0.05, "grad_norm": 1.377567172050476, "learning_rate": 9.948647843019705e-05, "loss": 2.4023, "step": 5880 }, { "epoch": 0.05, "grad_norm": 4.14026403427124, "learning_rate": 9.948559112751456e-05, "loss": 2.3621, "step": 5885 }, { "epoch": 0.05, "grad_norm": 4.878298759460449, "learning_rate": 9.948470306288189e-05, "loss": 2.4733, "step": 5890 }, { "epoch": 0.05, "grad_norm": 0.6675955653190613, "learning_rate": 9.948381423631273e-05, "loss": 2.5343, "step": 5895 }, { "epoch": 0.05, "grad_norm": 0.9237160086631775, "learning_rate": 9.948292464782072e-05, "loss": 2.4852, "step": 5900 }, { "epoch": 0.05, "grad_norm": 1.3330504894256592, "learning_rate": 9.948203429741961e-05, "loss": 2.3841, "step": 5905 }, { "epoch": 0.05, "grad_norm": 0.8018707036972046, "learning_rate": 9.948114318512307e-05, "loss": 2.4255, "step": 5910 }, { "epoch": 0.05, "grad_norm": 2.9265599250793457, "learning_rate": 9.948025131094482e-05, "loss": 2.3028, "step": 5915 }, { "epoch": 0.05, "grad_norm": 0.6129043698310852, "learning_rate": 9.947935867489862e-05, "loss": 2.3386, "step": 5920 }, { "epoch": 0.05, "grad_norm": 0.9805408120155334, "learning_rate": 9.94784652769982e-05, "loss": 2.4219, "step": 5925 }, { "epoch": 0.05, "grad_norm": 1.7506929636001587, "learning_rate": 9.947757111725732e-05, "loss": 2.5002, "step": 5930 }, { "epoch": 0.05, "grad_norm": 0.6212351322174072, "learning_rate": 9.947667619568974e-05, "loss": 2.401, "step": 5935 }, { "epoch": 0.05, "grad_norm": 0.6768795847892761, "learning_rate": 9.947578051230925e-05, "loss": 2.3852, "step": 5940 }, { "epoch": 0.05, "grad_norm": 0.8513750433921814, "learning_rate": 9.947488406712962e-05, "loss": 2.4523, "step": 5945 }, { "epoch": 0.05, "grad_norm": 0.7004748582839966, "learning_rate": 9.947398686016469e-05, "loss": 2.3344, "step": 5950 }, { "epoch": 0.05, "grad_norm": 1.0657085180282593, "learning_rate": 9.947308889142824e-05, "loss": 2.4328, "step": 5955 }, { "epoch": 0.05, "grad_norm": 0.6963376998901367, "learning_rate": 9.94721901609341e-05, "loss": 2.5145, "step": 5960 }, { "epoch": 0.05, "grad_norm": 0.8953424096107483, "learning_rate": 9.947129066869614e-05, "loss": 2.4033, "step": 5965 }, { "epoch": 0.05, "grad_norm": 0.7976968884468079, "learning_rate": 9.947039041472817e-05, "loss": 2.3372, "step": 5970 }, { "epoch": 0.05, "grad_norm": 0.8757538795471191, "learning_rate": 9.946948939904407e-05, "loss": 2.376, "step": 5975 }, { "epoch": 0.05, "grad_norm": 0.830988347530365, "learning_rate": 9.946858762165772e-05, "loss": 2.4034, "step": 5980 }, { "epoch": 0.05, "grad_norm": 1.6592557430267334, "learning_rate": 9.946768508258299e-05, "loss": 2.4825, "step": 5985 }, { "epoch": 0.05, "grad_norm": 0.5760574340820312, "learning_rate": 9.946678178183377e-05, "loss": 2.3425, "step": 5990 }, { "epoch": 0.05, "grad_norm": 0.5547943711280823, "learning_rate": 9.946587771942401e-05, "loss": 2.3918, "step": 5995 }, { "epoch": 0.05, "grad_norm": 0.6547987461090088, "learning_rate": 9.946497289536758e-05, "loss": 2.4185, "step": 6000 }, { "epoch": 0.05, "grad_norm": 1.2581392526626587, "learning_rate": 9.946406730967846e-05, "loss": 2.4731, "step": 6005 }, { "epoch": 0.05, "grad_norm": 3.9791390895843506, "learning_rate": 9.946316096237054e-05, "loss": 2.3932, "step": 6010 }, { "epoch": 0.05, "grad_norm": 0.6006248593330383, "learning_rate": 9.946225385345782e-05, "loss": 2.366, "step": 6015 }, { "epoch": 0.05, "grad_norm": 2.0880799293518066, "learning_rate": 9.946134598295424e-05, "loss": 2.3826, "step": 6020 }, { "epoch": 0.05, "grad_norm": 1.681903600692749, "learning_rate": 9.94604373508738e-05, "loss": 2.4089, "step": 6025 }, { "epoch": 0.05, "grad_norm": 0.5639763474464417, "learning_rate": 9.945952795723048e-05, "loss": 2.3959, "step": 6030 }, { "epoch": 0.05, "grad_norm": 0.8167706727981567, "learning_rate": 9.945861780203828e-05, "loss": 2.4327, "step": 6035 }, { "epoch": 0.05, "grad_norm": 0.88368821144104, "learning_rate": 9.945770688531122e-05, "loss": 2.4819, "step": 6040 }, { "epoch": 0.05, "grad_norm": 0.6680841445922852, "learning_rate": 9.945679520706332e-05, "loss": 2.4281, "step": 6045 }, { "epoch": 0.05, "grad_norm": 1.3033649921417236, "learning_rate": 9.945588276730862e-05, "loss": 2.3718, "step": 6050 }, { "epoch": 0.05, "grad_norm": 0.9796964526176453, "learning_rate": 9.945496956606115e-05, "loss": 2.4215, "step": 6055 }, { "epoch": 0.05, "grad_norm": 2.6386451721191406, "learning_rate": 9.945405560333502e-05, "loss": 2.3826, "step": 6060 }, { "epoch": 0.05, "grad_norm": 1.3074036836624146, "learning_rate": 9.945314087914426e-05, "loss": 2.4041, "step": 6065 }, { "epoch": 0.05, "grad_norm": 0.8867496252059937, "learning_rate": 9.945222539350296e-05, "loss": 2.3111, "step": 6070 }, { "epoch": 0.05, "grad_norm": 6.427149772644043, "learning_rate": 9.945130914642524e-05, "loss": 2.4478, "step": 6075 }, { "epoch": 0.05, "grad_norm": 1.0161908864974976, "learning_rate": 9.945039213792518e-05, "loss": 2.5012, "step": 6080 }, { "epoch": 0.05, "grad_norm": 4.425612449645996, "learning_rate": 9.94494743680169e-05, "loss": 2.4113, "step": 6085 }, { "epoch": 0.05, "grad_norm": 0.764665424823761, "learning_rate": 9.944855583671455e-05, "loss": 2.3896, "step": 6090 }, { "epoch": 0.05, "grad_norm": 10.818696975708008, "learning_rate": 9.944763654403226e-05, "loss": 2.4575, "step": 6095 }, { "epoch": 0.05, "grad_norm": 0.7486881017684937, "learning_rate": 9.94467164899842e-05, "loss": 2.3692, "step": 6100 }, { "epoch": 0.05, "grad_norm": 0.5912533402442932, "learning_rate": 9.944579567458452e-05, "loss": 2.4045, "step": 6105 }, { "epoch": 0.05, "grad_norm": 3.423297643661499, "learning_rate": 9.944487409784741e-05, "loss": 2.5357, "step": 6110 }, { "epoch": 0.05, "grad_norm": 1.105198860168457, "learning_rate": 9.944395175978705e-05, "loss": 2.4249, "step": 6115 }, { "epoch": 0.05, "grad_norm": 0.5690335035324097, "learning_rate": 9.944302866041764e-05, "loss": 2.4152, "step": 6120 }, { "epoch": 0.05, "grad_norm": 4.445101737976074, "learning_rate": 9.944210479975342e-05, "loss": 2.4502, "step": 6125 }, { "epoch": 0.05, "grad_norm": 0.7943013906478882, "learning_rate": 9.944118017780857e-05, "loss": 2.4458, "step": 6130 }, { "epoch": 0.05, "grad_norm": 1.4629900455474854, "learning_rate": 9.944025479459735e-05, "loss": 2.5338, "step": 6135 }, { "epoch": 0.05, "grad_norm": 8.919564247131348, "learning_rate": 9.943932865013404e-05, "loss": 2.4169, "step": 6140 }, { "epoch": 0.05, "grad_norm": 0.7106007933616638, "learning_rate": 9.943840174443284e-05, "loss": 2.3771, "step": 6145 }, { "epoch": 0.05, "grad_norm": 2.0090534687042236, "learning_rate": 9.943747407750805e-05, "loss": 2.4172, "step": 6150 }, { "epoch": 0.05, "grad_norm": 2.035435438156128, "learning_rate": 9.943654564937397e-05, "loss": 2.4226, "step": 6155 }, { "epoch": 0.05, "grad_norm": 0.6621341705322266, "learning_rate": 9.943561646004488e-05, "loss": 2.3314, "step": 6160 }, { "epoch": 0.05, "grad_norm": 0.7873374819755554, "learning_rate": 9.943468650953508e-05, "loss": 2.4487, "step": 6165 }, { "epoch": 0.05, "grad_norm": 0.6527106761932373, "learning_rate": 9.943375579785889e-05, "loss": 2.4607, "step": 6170 }, { "epoch": 0.05, "grad_norm": 0.5658855438232422, "learning_rate": 9.943282432503067e-05, "loss": 2.3303, "step": 6175 }, { "epoch": 0.05, "grad_norm": 3.661393404006958, "learning_rate": 9.943189209106472e-05, "loss": 2.3483, "step": 6180 }, { "epoch": 0.05, "grad_norm": 0.5517950057983398, "learning_rate": 9.943095909597541e-05, "loss": 2.3658, "step": 6185 }, { "epoch": 0.05, "grad_norm": 0.5192115306854248, "learning_rate": 9.943002533977713e-05, "loss": 2.356, "step": 6190 }, { "epoch": 0.05, "grad_norm": 3.115165948867798, "learning_rate": 9.942909082248423e-05, "loss": 2.4314, "step": 6195 }, { "epoch": 0.05, "grad_norm": 1.0805926322937012, "learning_rate": 9.942815554411109e-05, "loss": 2.4169, "step": 6200 }, { "epoch": 0.05, "grad_norm": 1.2299894094467163, "learning_rate": 9.942721950467214e-05, "loss": 2.4428, "step": 6205 }, { "epoch": 0.05, "grad_norm": 0.9934828281402588, "learning_rate": 9.942628270418177e-05, "loss": 2.3384, "step": 6210 }, { "epoch": 0.05, "grad_norm": 0.9799348711967468, "learning_rate": 9.942534514265442e-05, "loss": 2.3604, "step": 6215 }, { "epoch": 0.05, "grad_norm": 3.160379648208618, "learning_rate": 9.942440682010452e-05, "loss": 2.4841, "step": 6220 }, { "epoch": 0.05, "grad_norm": 1.0762101411819458, "learning_rate": 9.942346773654652e-05, "loss": 2.3792, "step": 6225 }, { "epoch": 0.05, "grad_norm": 0.5418878793716431, "learning_rate": 9.942252789199487e-05, "loss": 2.4405, "step": 6230 }, { "epoch": 0.05, "grad_norm": 12.935803413391113, "learning_rate": 9.942158728646406e-05, "loss": 2.4772, "step": 6235 }, { "epoch": 0.05, "grad_norm": 1.023612380027771, "learning_rate": 9.942064591996855e-05, "loss": 2.4471, "step": 6240 }, { "epoch": 0.05, "grad_norm": 2.3708746433258057, "learning_rate": 9.941970379252284e-05, "loss": 2.3892, "step": 6245 }, { "epoch": 0.05, "grad_norm": 1.0352396965026855, "learning_rate": 9.941876090414146e-05, "loss": 2.3428, "step": 6250 }, { "epoch": 0.05, "grad_norm": 5.042909145355225, "learning_rate": 9.941781725483889e-05, "loss": 2.5418, "step": 6255 }, { "epoch": 0.05, "grad_norm": 3.395291566848755, "learning_rate": 9.941687284462969e-05, "loss": 2.4332, "step": 6260 }, { "epoch": 0.05, "grad_norm": 0.7150275111198425, "learning_rate": 9.941592767352839e-05, "loss": 2.4575, "step": 6265 }, { "epoch": 0.05, "grad_norm": 0.5269148349761963, "learning_rate": 9.941498174154954e-05, "loss": 2.3747, "step": 6270 }, { "epoch": 0.05, "grad_norm": 0.7054476141929626, "learning_rate": 9.94140350487077e-05, "loss": 2.4019, "step": 6275 }, { "epoch": 0.05, "grad_norm": 0.6725271940231323, "learning_rate": 9.941308759501747e-05, "loss": 2.4222, "step": 6280 }, { "epoch": 0.05, "grad_norm": 2.6249637603759766, "learning_rate": 9.941213938049342e-05, "loss": 2.447, "step": 6285 }, { "epoch": 0.05, "grad_norm": 1.0070902109146118, "learning_rate": 9.941119040515017e-05, "loss": 2.3791, "step": 6290 }, { "epoch": 0.05, "grad_norm": 0.5993001461029053, "learning_rate": 9.941024066900229e-05, "loss": 2.441, "step": 6295 }, { "epoch": 0.05, "grad_norm": 1.3684808015823364, "learning_rate": 9.940929017206445e-05, "loss": 2.4087, "step": 6300 }, { "epoch": 0.05, "grad_norm": 0.6140982508659363, "learning_rate": 9.940833891435124e-05, "loss": 2.3987, "step": 6305 }, { "epoch": 0.05, "grad_norm": 10.819245338439941, "learning_rate": 9.940738689587735e-05, "loss": 2.3691, "step": 6310 }, { "epoch": 0.05, "grad_norm": 0.8927221298217773, "learning_rate": 9.940643411665742e-05, "loss": 2.3906, "step": 6315 }, { "epoch": 0.05, "grad_norm": 15.180883407592773, "learning_rate": 9.940548057670612e-05, "loss": 2.5625, "step": 6320 }, { "epoch": 0.05, "grad_norm": 1.6519043445587158, "learning_rate": 9.940452627603813e-05, "loss": 2.3884, "step": 6325 }, { "epoch": 0.05, "grad_norm": 0.6707853078842163, "learning_rate": 9.940357121466813e-05, "loss": 2.3955, "step": 6330 }, { "epoch": 0.05, "grad_norm": 1.162865400314331, "learning_rate": 9.940261539261086e-05, "loss": 2.4119, "step": 6335 }, { "epoch": 0.05, "grad_norm": 1.2904102802276611, "learning_rate": 9.940165880988102e-05, "loss": 2.4038, "step": 6340 }, { "epoch": 0.05, "grad_norm": 0.9950482249259949, "learning_rate": 9.940070146649333e-05, "loss": 2.3581, "step": 6345 }, { "epoch": 0.05, "grad_norm": 13.598631858825684, "learning_rate": 9.939974336246254e-05, "loss": 2.4218, "step": 6350 }, { "epoch": 0.05, "grad_norm": 0.6584912538528442, "learning_rate": 9.93987844978034e-05, "loss": 2.4764, "step": 6355 }, { "epoch": 0.05, "grad_norm": 0.5133340358734131, "learning_rate": 9.939782487253067e-05, "loss": 2.4639, "step": 6360 }, { "epoch": 0.05, "grad_norm": 5.230878829956055, "learning_rate": 9.939686448665915e-05, "loss": 2.3568, "step": 6365 }, { "epoch": 0.05, "grad_norm": 1.2003432512283325, "learning_rate": 9.939590334020359e-05, "loss": 2.5034, "step": 6370 }, { "epoch": 0.05, "grad_norm": 2.2117743492126465, "learning_rate": 9.939494143317881e-05, "loss": 2.429, "step": 6375 }, { "epoch": 0.05, "grad_norm": 0.6651952862739563, "learning_rate": 9.939397876559962e-05, "loss": 2.3975, "step": 6380 }, { "epoch": 0.05, "grad_norm": 0.8760754466056824, "learning_rate": 9.939301533748084e-05, "loss": 2.4395, "step": 6385 }, { "epoch": 0.05, "grad_norm": 0.514520525932312, "learning_rate": 9.939205114883731e-05, "loss": 2.3786, "step": 6390 }, { "epoch": 0.05, "grad_norm": 7.4580583572387695, "learning_rate": 9.939108619968386e-05, "loss": 2.4059, "step": 6395 }, { "epoch": 0.05, "grad_norm": 1.80430006980896, "learning_rate": 9.939012049003537e-05, "loss": 2.4585, "step": 6400 }, { "epoch": 0.05, "grad_norm": 3.011021375656128, "learning_rate": 9.93891540199067e-05, "loss": 2.3973, "step": 6405 }, { "epoch": 0.05, "grad_norm": 0.5773200392723083, "learning_rate": 9.938818678931271e-05, "loss": 2.3798, "step": 6410 }, { "epoch": 0.05, "grad_norm": 0.8097543716430664, "learning_rate": 9.938721879826832e-05, "loss": 2.2787, "step": 6415 }, { "epoch": 0.05, "grad_norm": 2.2829787731170654, "learning_rate": 9.938625004678844e-05, "loss": 2.3808, "step": 6420 }, { "epoch": 0.05, "grad_norm": 1.2852363586425781, "learning_rate": 9.938528053488796e-05, "loss": 2.4464, "step": 6425 }, { "epoch": 0.05, "grad_norm": 1.0180230140686035, "learning_rate": 9.938431026258182e-05, "loss": 2.4376, "step": 6430 }, { "epoch": 0.05, "grad_norm": 0.6233970522880554, "learning_rate": 9.938333922988497e-05, "loss": 2.3926, "step": 6435 }, { "epoch": 0.05, "grad_norm": 7.141419887542725, "learning_rate": 9.938236743681235e-05, "loss": 2.3958, "step": 6440 }, { "epoch": 0.05, "grad_norm": 1.465430498123169, "learning_rate": 9.938139488337891e-05, "loss": 2.4112, "step": 6445 }, { "epoch": 0.05, "grad_norm": 2.3740384578704834, "learning_rate": 9.938042156959966e-05, "loss": 2.3715, "step": 6450 }, { "epoch": 0.05, "grad_norm": 0.6643574833869934, "learning_rate": 9.937944749548956e-05, "loss": 2.5248, "step": 6455 }, { "epoch": 0.05, "grad_norm": 2.1322879791259766, "learning_rate": 9.937847266106359e-05, "loss": 2.4417, "step": 6460 }, { "epoch": 0.05, "grad_norm": 1.3981804847717285, "learning_rate": 9.937749706633682e-05, "loss": 2.3884, "step": 6465 }, { "epoch": 0.05, "grad_norm": 0.7077898383140564, "learning_rate": 9.937652071132422e-05, "loss": 2.4642, "step": 6470 }, { "epoch": 0.05, "grad_norm": 0.5573113560676575, "learning_rate": 9.937554359604084e-05, "loss": 2.373, "step": 6475 }, { "epoch": 0.05, "grad_norm": 0.868512749671936, "learning_rate": 9.937456572050172e-05, "loss": 2.398, "step": 6480 }, { "epoch": 0.05, "grad_norm": 1.0273706912994385, "learning_rate": 9.937358708472193e-05, "loss": 2.4282, "step": 6485 }, { "epoch": 0.05, "grad_norm": 0.8703930974006653, "learning_rate": 9.937260768871654e-05, "loss": 2.4295, "step": 6490 }, { "epoch": 0.05, "grad_norm": 1.1538885831832886, "learning_rate": 9.93716275325006e-05, "loss": 2.3667, "step": 6495 }, { "epoch": 0.05, "grad_norm": 1.4038690328598022, "learning_rate": 9.937064661608924e-05, "loss": 2.3395, "step": 6500 }, { "epoch": 0.05, "grad_norm": 1.0305116176605225, "learning_rate": 9.936966493949752e-05, "loss": 2.3929, "step": 6505 }, { "epoch": 0.05, "grad_norm": 1.0581971406936646, "learning_rate": 9.936868250274059e-05, "loss": 2.388, "step": 6510 }, { "epoch": 0.05, "grad_norm": 1.2772094011306763, "learning_rate": 9.936769930583359e-05, "loss": 2.364, "step": 6515 }, { "epoch": 0.05, "grad_norm": 0.7324458956718445, "learning_rate": 9.93667153487916e-05, "loss": 2.3207, "step": 6520 }, { "epoch": 0.05, "grad_norm": 38.07729721069336, "learning_rate": 9.936573063162984e-05, "loss": 2.4231, "step": 6525 }, { "epoch": 0.05, "grad_norm": 0.9465434551239014, "learning_rate": 9.936474515436342e-05, "loss": 2.3836, "step": 6530 }, { "epoch": 0.05, "grad_norm": 0.7391358017921448, "learning_rate": 9.936375891700753e-05, "loss": 2.4487, "step": 6535 }, { "epoch": 0.05, "grad_norm": 0.901811957359314, "learning_rate": 9.936277191957737e-05, "loss": 2.4603, "step": 6540 }, { "epoch": 0.05, "grad_norm": 1.2855156660079956, "learning_rate": 9.936178416208813e-05, "loss": 2.3464, "step": 6545 }, { "epoch": 0.05, "grad_norm": 4.146651744842529, "learning_rate": 9.9360795644555e-05, "loss": 2.3983, "step": 6550 }, { "epoch": 0.05, "grad_norm": 1.6682440042495728, "learning_rate": 9.935980636699322e-05, "loss": 2.4568, "step": 6555 }, { "epoch": 0.05, "grad_norm": 1.157954454421997, "learning_rate": 9.935881632941801e-05, "loss": 2.5005, "step": 6560 }, { "epoch": 0.05, "grad_norm": 0.6005189418792725, "learning_rate": 9.935782553184463e-05, "loss": 2.2753, "step": 6565 }, { "epoch": 0.05, "grad_norm": 0.9284230470657349, "learning_rate": 9.935683397428833e-05, "loss": 2.4498, "step": 6570 }, { "epoch": 0.05, "grad_norm": 7.745828628540039, "learning_rate": 9.935584165676436e-05, "loss": 2.4271, "step": 6575 }, { "epoch": 0.05, "grad_norm": 1.0953933000564575, "learning_rate": 9.935484857928801e-05, "loss": 2.3803, "step": 6580 }, { "epoch": 0.05, "grad_norm": 0.6444041728973389, "learning_rate": 9.935385474187459e-05, "loss": 2.3937, "step": 6585 }, { "epoch": 0.05, "grad_norm": 3.5930566787719727, "learning_rate": 9.93528601445394e-05, "loss": 2.4391, "step": 6590 }, { "epoch": 0.05, "grad_norm": 0.5998481512069702, "learning_rate": 9.935186478729771e-05, "loss": 2.3867, "step": 6595 }, { "epoch": 0.05, "grad_norm": 1.4485433101654053, "learning_rate": 9.93508686701649e-05, "loss": 2.4136, "step": 6600 }, { "epoch": 0.05, "grad_norm": 0.5119336843490601, "learning_rate": 9.934987179315627e-05, "loss": 2.3545, "step": 6605 }, { "epoch": 0.05, "grad_norm": 0.6237988471984863, "learning_rate": 9.934887415628719e-05, "loss": 2.3157, "step": 6610 }, { "epoch": 0.05, "grad_norm": 0.8357030749320984, "learning_rate": 9.934787575957301e-05, "loss": 2.3832, "step": 6615 }, { "epoch": 0.05, "grad_norm": 1.15729820728302, "learning_rate": 9.934687660302912e-05, "loss": 2.4097, "step": 6620 }, { "epoch": 0.05, "grad_norm": 0.5791693925857544, "learning_rate": 9.934587668667088e-05, "loss": 2.4278, "step": 6625 }, { "epoch": 0.05, "grad_norm": 0.5358119010925293, "learning_rate": 9.93448760105137e-05, "loss": 2.3711, "step": 6630 }, { "epoch": 0.05, "grad_norm": 0.7729812860488892, "learning_rate": 9.934387457457299e-05, "loss": 2.4037, "step": 6635 }, { "epoch": 0.05, "grad_norm": 13.806668281555176, "learning_rate": 9.934287237886417e-05, "loss": 2.431, "step": 6640 }, { "epoch": 0.05, "grad_norm": 1.0619170665740967, "learning_rate": 9.934186942340265e-05, "loss": 2.3199, "step": 6645 }, { "epoch": 0.05, "grad_norm": 3.9093427658081055, "learning_rate": 9.934086570820392e-05, "loss": 2.3191, "step": 6650 }, { "epoch": 0.05, "grad_norm": 0.7148905396461487, "learning_rate": 9.933986123328337e-05, "loss": 2.4156, "step": 6655 }, { "epoch": 0.05, "grad_norm": 3.5102427005767822, "learning_rate": 9.933885599865651e-05, "loss": 2.3933, "step": 6660 }, { "epoch": 0.05, "grad_norm": 0.7663382887840271, "learning_rate": 9.933785000433882e-05, "loss": 2.347, "step": 6665 }, { "epoch": 0.05, "grad_norm": 1.2696189880371094, "learning_rate": 9.933684325034577e-05, "loss": 2.426, "step": 6670 }, { "epoch": 0.05, "grad_norm": 0.6777164936065674, "learning_rate": 9.933583573669287e-05, "loss": 2.4434, "step": 6675 }, { "epoch": 0.05, "grad_norm": 6.8198466300964355, "learning_rate": 9.933482746339564e-05, "loss": 2.3249, "step": 6680 }, { "epoch": 0.05, "grad_norm": 1.1049549579620361, "learning_rate": 9.933381843046959e-05, "loss": 2.3716, "step": 6685 }, { "epoch": 0.05, "grad_norm": 0.8847483992576599, "learning_rate": 9.933280863793026e-05, "loss": 2.4178, "step": 6690 }, { "epoch": 0.05, "grad_norm": 6.436984062194824, "learning_rate": 9.933179808579321e-05, "loss": 2.3843, "step": 6695 }, { "epoch": 0.05, "grad_norm": 0.8163633942604065, "learning_rate": 9.933078677407399e-05, "loss": 2.3905, "step": 6700 }, { "epoch": 0.05, "grad_norm": 0.7351194024085999, "learning_rate": 9.932977470278818e-05, "loss": 2.3683, "step": 6705 }, { "epoch": 0.05, "grad_norm": 5.296935558319092, "learning_rate": 9.932876187195135e-05, "loss": 2.3887, "step": 6710 }, { "epoch": 0.05, "grad_norm": 0.5954237580299377, "learning_rate": 9.932774828157911e-05, "loss": 2.3376, "step": 6715 }, { "epoch": 0.05, "grad_norm": 1.1438419818878174, "learning_rate": 9.932673393168705e-05, "loss": 2.5356, "step": 6720 }, { "epoch": 0.05, "grad_norm": 0.6553308367729187, "learning_rate": 9.932571882229081e-05, "loss": 2.4471, "step": 6725 }, { "epoch": 0.05, "grad_norm": 1.6205651760101318, "learning_rate": 9.932470295340599e-05, "loss": 2.2925, "step": 6730 }, { "epoch": 0.05, "grad_norm": 0.8698533773422241, "learning_rate": 9.932368632504826e-05, "loss": 2.3141, "step": 6735 }, { "epoch": 0.05, "grad_norm": 0.5471912026405334, "learning_rate": 9.932266893723325e-05, "loss": 2.4062, "step": 6740 }, { "epoch": 0.05, "grad_norm": 1.4634628295898438, "learning_rate": 9.932165078997666e-05, "loss": 2.4755, "step": 6745 }, { "epoch": 0.05, "grad_norm": 0.8293861746788025, "learning_rate": 9.932063188329412e-05, "loss": 2.3965, "step": 6750 }, { "epoch": 0.05, "grad_norm": 0.8285143971443176, "learning_rate": 9.931961221720136e-05, "loss": 2.3412, "step": 6755 }, { "epoch": 0.05, "grad_norm": 0.9669542908668518, "learning_rate": 9.931859179171405e-05, "loss": 2.3245, "step": 6760 }, { "epoch": 0.05, "grad_norm": 1.0269503593444824, "learning_rate": 9.931757060684792e-05, "loss": 2.4238, "step": 6765 }, { "epoch": 0.05, "grad_norm": 0.8528624176979065, "learning_rate": 9.93165486626187e-05, "loss": 2.4126, "step": 6770 }, { "epoch": 0.05, "grad_norm": 1.235798954963684, "learning_rate": 9.93155259590421e-05, "loss": 2.3722, "step": 6775 }, { "epoch": 0.05, "grad_norm": 1.816422939300537, "learning_rate": 9.931450249613389e-05, "loss": 2.3469, "step": 6780 }, { "epoch": 0.05, "grad_norm": 1.1803120374679565, "learning_rate": 9.931347827390982e-05, "loss": 2.4194, "step": 6785 }, { "epoch": 0.05, "grad_norm": 1.8255891799926758, "learning_rate": 9.931245329238566e-05, "loss": 2.4437, "step": 6790 }, { "epoch": 0.05, "grad_norm": 1.02006995677948, "learning_rate": 9.931142755157719e-05, "loss": 2.3941, "step": 6795 }, { "epoch": 0.05, "grad_norm": 3.196514368057251, "learning_rate": 9.931040105150019e-05, "loss": 2.446, "step": 6800 }, { "epoch": 0.05, "grad_norm": 0.5649253129959106, "learning_rate": 9.93093737921705e-05, "loss": 2.3292, "step": 6805 }, { "epoch": 0.05, "grad_norm": 10.325676918029785, "learning_rate": 9.93083457736039e-05, "loss": 2.5524, "step": 6810 }, { "epoch": 0.05, "grad_norm": 0.8547642827033997, "learning_rate": 9.930731699581626e-05, "loss": 2.4121, "step": 6815 }, { "epoch": 0.05, "grad_norm": 3.472689390182495, "learning_rate": 9.930628745882339e-05, "loss": 2.4089, "step": 6820 }, { "epoch": 0.05, "grad_norm": 1.327664852142334, "learning_rate": 9.930525716264114e-05, "loss": 2.4295, "step": 6825 }, { "epoch": 0.05, "grad_norm": 5.6567535400390625, "learning_rate": 9.93042261072854e-05, "loss": 2.3831, "step": 6830 }, { "epoch": 0.05, "grad_norm": 0.6467896103858948, "learning_rate": 9.930319429277202e-05, "loss": 2.3905, "step": 6835 }, { "epoch": 0.05, "grad_norm": 0.5680961608886719, "learning_rate": 9.93021617191169e-05, "loss": 2.4172, "step": 6840 }, { "epoch": 0.05, "grad_norm": 0.8354523777961731, "learning_rate": 9.930112838633592e-05, "loss": 2.3047, "step": 6845 }, { "epoch": 0.05, "grad_norm": 0.7659814953804016, "learning_rate": 9.930009429444503e-05, "loss": 2.3748, "step": 6850 }, { "epoch": 0.05, "grad_norm": 14.04178524017334, "learning_rate": 9.929905944346012e-05, "loss": 2.2736, "step": 6855 }, { "epoch": 0.05, "grad_norm": 0.6975845694541931, "learning_rate": 9.929802383339713e-05, "loss": 2.4377, "step": 6860 }, { "epoch": 0.05, "grad_norm": 1.1097325086593628, "learning_rate": 9.9296987464272e-05, "loss": 2.5002, "step": 6865 }, { "epoch": 0.05, "grad_norm": 7.024208068847656, "learning_rate": 9.929595033610072e-05, "loss": 2.3621, "step": 6870 }, { "epoch": 0.05, "grad_norm": 0.5781006217002869, "learning_rate": 9.929491244889921e-05, "loss": 2.4485, "step": 6875 }, { "epoch": 0.05, "grad_norm": 0.9415589570999146, "learning_rate": 9.929387380268349e-05, "loss": 2.2666, "step": 6880 }, { "epoch": 0.05, "grad_norm": 0.8640931844711304, "learning_rate": 9.929283439746955e-05, "loss": 2.387, "step": 6885 }, { "epoch": 0.05, "grad_norm": 1.026068091392517, "learning_rate": 9.929179423327336e-05, "loss": 2.3938, "step": 6890 }, { "epoch": 0.05, "grad_norm": 0.9859316945075989, "learning_rate": 9.929075331011097e-05, "loss": 2.3691, "step": 6895 }, { "epoch": 0.05, "grad_norm": 0.79823899269104, "learning_rate": 9.928971162799838e-05, "loss": 2.393, "step": 6900 }, { "epoch": 0.05, "grad_norm": 1.0991322994232178, "learning_rate": 9.928866918695165e-05, "loss": 2.3157, "step": 6905 }, { "epoch": 0.05, "grad_norm": 0.930264949798584, "learning_rate": 9.928762598698684e-05, "loss": 2.4742, "step": 6910 }, { "epoch": 0.05, "grad_norm": 1.3068737983703613, "learning_rate": 9.928658202811999e-05, "loss": 2.3273, "step": 6915 }, { "epoch": 0.05, "grad_norm": 0.7164826989173889, "learning_rate": 9.928553731036719e-05, "loss": 2.4314, "step": 6920 }, { "epoch": 0.05, "grad_norm": 1.5798934698104858, "learning_rate": 9.928449183374451e-05, "loss": 2.4285, "step": 6925 }, { "epoch": 0.05, "grad_norm": 1.0284011363983154, "learning_rate": 9.928344559826806e-05, "loss": 2.3839, "step": 6930 }, { "epoch": 0.05, "grad_norm": 0.7840524911880493, "learning_rate": 9.928239860395394e-05, "loss": 2.4158, "step": 6935 }, { "epoch": 0.05, "grad_norm": 1.3443907499313354, "learning_rate": 9.928135085081827e-05, "loss": 2.4356, "step": 6940 }, { "epoch": 0.05, "grad_norm": 0.5811484456062317, "learning_rate": 9.928030233887721e-05, "loss": 2.376, "step": 6945 }, { "epoch": 0.05, "grad_norm": 1.1974537372589111, "learning_rate": 9.927925306814686e-05, "loss": 2.4399, "step": 6950 }, { "epoch": 0.05, "grad_norm": 0.8177313804626465, "learning_rate": 9.927820303864341e-05, "loss": 2.3514, "step": 6955 }, { "epoch": 0.05, "grad_norm": 1.1507309675216675, "learning_rate": 9.927715225038303e-05, "loss": 2.3658, "step": 6960 }, { "epoch": 0.05, "grad_norm": 1.1967442035675049, "learning_rate": 9.927610070338187e-05, "loss": 2.4207, "step": 6965 }, { "epoch": 0.06, "grad_norm": 0.6558987498283386, "learning_rate": 9.927504839765615e-05, "loss": 2.4627, "step": 6970 }, { "epoch": 0.06, "grad_norm": 1.4211279153823853, "learning_rate": 9.927399533322206e-05, "loss": 2.3223, "step": 6975 }, { "epoch": 0.06, "grad_norm": 0.6763742566108704, "learning_rate": 9.92729415100958e-05, "loss": 2.3499, "step": 6980 }, { "epoch": 0.06, "grad_norm": 0.6545825004577637, "learning_rate": 9.927188692829363e-05, "loss": 2.3899, "step": 6985 }, { "epoch": 0.06, "grad_norm": 0.5871744155883789, "learning_rate": 9.927083158783177e-05, "loss": 2.3783, "step": 6990 }, { "epoch": 0.06, "grad_norm": 7.306336402893066, "learning_rate": 9.926977548872644e-05, "loss": 2.4225, "step": 6995 }, { "epoch": 0.06, "grad_norm": 1.0360629558563232, "learning_rate": 9.926871863099396e-05, "loss": 2.4264, "step": 7000 }, { "epoch": 0.06, "grad_norm": 32.41043472290039, "learning_rate": 9.926766101465056e-05, "loss": 2.5008, "step": 7005 }, { "epoch": 0.06, "grad_norm": 0.6528191566467285, "learning_rate": 9.926660263971254e-05, "loss": 2.394, "step": 7010 }, { "epoch": 0.06, "grad_norm": 1.510021686553955, "learning_rate": 9.92655435061962e-05, "loss": 2.3535, "step": 7015 }, { "epoch": 0.06, "grad_norm": 1.1839497089385986, "learning_rate": 9.926448361411784e-05, "loss": 2.4438, "step": 7020 }, { "epoch": 0.06, "grad_norm": 0.9368022680282593, "learning_rate": 9.926342296349378e-05, "loss": 2.3749, "step": 7025 }, { "epoch": 0.06, "grad_norm": 1.3842153549194336, "learning_rate": 9.926236155434034e-05, "loss": 2.4461, "step": 7030 }, { "epoch": 0.06, "grad_norm": 0.6904487609863281, "learning_rate": 9.926129938667388e-05, "loss": 2.432, "step": 7035 }, { "epoch": 0.06, "grad_norm": 11.606537818908691, "learning_rate": 9.926023646051075e-05, "loss": 2.402, "step": 7040 }, { "epoch": 0.06, "grad_norm": 1.3381249904632568, "learning_rate": 9.925917277586734e-05, "loss": 2.3752, "step": 7045 }, { "epoch": 0.06, "grad_norm": 0.8679342269897461, "learning_rate": 9.925810833275998e-05, "loss": 2.3868, "step": 7050 }, { "epoch": 0.06, "grad_norm": 0.6749829649925232, "learning_rate": 9.925704313120508e-05, "loss": 2.3346, "step": 7055 }, { "epoch": 0.06, "grad_norm": 0.6255229115486145, "learning_rate": 9.925597717121906e-05, "loss": 2.4108, "step": 7060 }, { "epoch": 0.06, "grad_norm": 0.7203311324119568, "learning_rate": 9.925491045281832e-05, "loss": 2.3909, "step": 7065 }, { "epoch": 0.06, "grad_norm": 2.665862560272217, "learning_rate": 9.925384297601928e-05, "loss": 2.3583, "step": 7070 }, { "epoch": 0.06, "grad_norm": 0.9749100208282471, "learning_rate": 9.925277474083835e-05, "loss": 2.3871, "step": 7075 }, { "epoch": 0.06, "grad_norm": 0.843224287033081, "learning_rate": 9.925170574729204e-05, "loss": 2.4579, "step": 7080 }, { "epoch": 0.06, "grad_norm": 3.1739230155944824, "learning_rate": 9.925063599539677e-05, "loss": 2.3609, "step": 7085 }, { "epoch": 0.06, "grad_norm": 0.669441282749176, "learning_rate": 9.924956548516901e-05, "loss": 2.447, "step": 7090 }, { "epoch": 0.06, "grad_norm": 0.9815073013305664, "learning_rate": 9.924849421662528e-05, "loss": 2.3529, "step": 7095 }, { "epoch": 0.06, "grad_norm": 31.384504318237305, "learning_rate": 9.924742218978203e-05, "loss": 2.3895, "step": 7100 }, { "epoch": 0.06, "grad_norm": 0.8862005472183228, "learning_rate": 9.924634940465577e-05, "loss": 2.4107, "step": 7105 }, { "epoch": 0.06, "grad_norm": 0.759222149848938, "learning_rate": 9.924527586126305e-05, "loss": 2.4782, "step": 7110 }, { "epoch": 0.06, "grad_norm": 0.8119022846221924, "learning_rate": 9.924420155962037e-05, "loss": 2.3431, "step": 7115 }, { "epoch": 0.06, "grad_norm": 0.7126891613006592, "learning_rate": 9.924312649974427e-05, "loss": 2.3748, "step": 7120 }, { "epoch": 0.06, "grad_norm": 2.4088804721832275, "learning_rate": 9.924205068165134e-05, "loss": 2.3728, "step": 7125 }, { "epoch": 0.06, "grad_norm": 0.9981231689453125, "learning_rate": 9.924097410535811e-05, "loss": 2.3993, "step": 7130 }, { "epoch": 0.06, "grad_norm": 1.5982328653335571, "learning_rate": 9.923989677088116e-05, "loss": 2.3856, "step": 7135 }, { "epoch": 0.06, "grad_norm": 1.3289297819137573, "learning_rate": 9.92388186782371e-05, "loss": 2.4336, "step": 7140 }, { "epoch": 0.06, "grad_norm": 14.112395286560059, "learning_rate": 9.92377398274425e-05, "loss": 2.3768, "step": 7145 }, { "epoch": 0.06, "grad_norm": 0.6084874868392944, "learning_rate": 9.923666021851398e-05, "loss": 2.3257, "step": 7150 }, { "epoch": 0.06, "grad_norm": 0.5459027886390686, "learning_rate": 9.92355798514682e-05, "loss": 2.4489, "step": 7155 }, { "epoch": 0.06, "grad_norm": 0.7760525345802307, "learning_rate": 9.923449872632173e-05, "loss": 2.3955, "step": 7160 }, { "epoch": 0.06, "grad_norm": 6.001205921173096, "learning_rate": 9.923341684309128e-05, "loss": 2.3211, "step": 7165 }, { "epoch": 0.06, "grad_norm": 0.7180355787277222, "learning_rate": 9.923233420179347e-05, "loss": 2.4528, "step": 7170 }, { "epoch": 0.06, "grad_norm": 0.8799859285354614, "learning_rate": 9.923125080244497e-05, "loss": 2.4609, "step": 7175 }, { "epoch": 0.06, "grad_norm": 9.965468406677246, "learning_rate": 9.923016664506248e-05, "loss": 2.2702, "step": 7180 }, { "epoch": 0.06, "grad_norm": 1.2095704078674316, "learning_rate": 9.922908172966268e-05, "loss": 2.4354, "step": 7185 }, { "epoch": 0.06, "grad_norm": 1.9123395681381226, "learning_rate": 9.922799605626227e-05, "loss": 2.3477, "step": 7190 }, { "epoch": 0.06, "grad_norm": 0.8337376117706299, "learning_rate": 9.9226909624878e-05, "loss": 2.3045, "step": 7195 }, { "epoch": 0.06, "grad_norm": 0.8681227564811707, "learning_rate": 9.922582243552654e-05, "loss": 2.4493, "step": 7200 }, { "epoch": 0.06, "grad_norm": 3.4228973388671875, "learning_rate": 9.922473448822469e-05, "loss": 2.381, "step": 7205 }, { "epoch": 0.06, "grad_norm": 0.8948184847831726, "learning_rate": 9.922364578298916e-05, "loss": 2.4474, "step": 7210 }, { "epoch": 0.06, "grad_norm": 0.7658812403678894, "learning_rate": 9.922255631983673e-05, "loss": 2.3627, "step": 7215 }, { "epoch": 0.06, "grad_norm": 6.623570919036865, "learning_rate": 9.922146609878418e-05, "loss": 2.3479, "step": 7220 }, { "epoch": 0.06, "grad_norm": 0.5731838941574097, "learning_rate": 9.922037511984829e-05, "loss": 2.3448, "step": 7225 }, { "epoch": 0.06, "grad_norm": 0.6682723760604858, "learning_rate": 9.921928338304584e-05, "loss": 2.4636, "step": 7230 }, { "epoch": 0.06, "grad_norm": 2.4286837577819824, "learning_rate": 9.921819088839369e-05, "loss": 2.4487, "step": 7235 }, { "epoch": 0.06, "grad_norm": 0.7052165269851685, "learning_rate": 9.92170976359086e-05, "loss": 2.4865, "step": 7240 }, { "epoch": 0.06, "grad_norm": 0.9136790037155151, "learning_rate": 9.921600362560744e-05, "loss": 2.3359, "step": 7245 }, { "epoch": 0.06, "grad_norm": 1.7511136531829834, "learning_rate": 9.921490885750706e-05, "loss": 2.3377, "step": 7250 }, { "epoch": 0.06, "grad_norm": 0.6979846358299255, "learning_rate": 9.92138133316243e-05, "loss": 2.3569, "step": 7255 }, { "epoch": 0.06, "grad_norm": 0.871650755405426, "learning_rate": 9.921271704797602e-05, "loss": 2.4022, "step": 7260 }, { "epoch": 0.06, "grad_norm": 0.6131813526153564, "learning_rate": 9.921162000657913e-05, "loss": 2.3556, "step": 7265 }, { "epoch": 0.06, "grad_norm": 1.3522052764892578, "learning_rate": 9.921052220745048e-05, "loss": 2.2949, "step": 7270 }, { "epoch": 0.06, "grad_norm": 0.616355836391449, "learning_rate": 9.920942365060702e-05, "loss": 2.3391, "step": 7275 }, { "epoch": 0.06, "grad_norm": 1.0883160829544067, "learning_rate": 9.920832433606564e-05, "loss": 2.3553, "step": 7280 }, { "epoch": 0.06, "grad_norm": 1.8716261386871338, "learning_rate": 9.920722426384326e-05, "loss": 2.325, "step": 7285 }, { "epoch": 0.06, "grad_norm": 1.7922699451446533, "learning_rate": 9.920612343395683e-05, "loss": 2.4536, "step": 7290 }, { "epoch": 0.06, "grad_norm": 0.9377602934837341, "learning_rate": 9.92050218464233e-05, "loss": 2.323, "step": 7295 }, { "epoch": 0.06, "grad_norm": 0.8790144920349121, "learning_rate": 9.920391950125963e-05, "loss": 2.3358, "step": 7300 }, { "epoch": 0.06, "grad_norm": 0.6641685962677002, "learning_rate": 9.920281639848278e-05, "loss": 2.3995, "step": 7305 }, { "epoch": 0.06, "grad_norm": 1.9648066759109497, "learning_rate": 9.920171253810975e-05, "loss": 2.4341, "step": 7310 }, { "epoch": 0.06, "grad_norm": 1.0014420747756958, "learning_rate": 9.920060792015754e-05, "loss": 2.3827, "step": 7315 }, { "epoch": 0.06, "grad_norm": 1.032088279724121, "learning_rate": 9.919950254464316e-05, "loss": 2.3822, "step": 7320 }, { "epoch": 0.06, "grad_norm": 0.741034209728241, "learning_rate": 9.919839641158361e-05, "loss": 2.4881, "step": 7325 }, { "epoch": 0.06, "grad_norm": 0.6413688659667969, "learning_rate": 9.919728952099593e-05, "loss": 2.391, "step": 7330 }, { "epoch": 0.06, "grad_norm": 1.2092832326889038, "learning_rate": 9.919618187289717e-05, "loss": 2.3543, "step": 7335 }, { "epoch": 0.06, "grad_norm": 0.8328579664230347, "learning_rate": 9.919507346730437e-05, "loss": 2.4584, "step": 7340 }, { "epoch": 0.06, "grad_norm": 0.6228640079498291, "learning_rate": 9.919396430423463e-05, "loss": 2.3724, "step": 7345 }, { "epoch": 0.06, "grad_norm": 1.6994060277938843, "learning_rate": 9.9192854383705e-05, "loss": 2.3813, "step": 7350 }, { "epoch": 0.06, "grad_norm": 1.1023279428482056, "learning_rate": 9.919174370573257e-05, "loss": 2.3474, "step": 7355 }, { "epoch": 0.06, "grad_norm": 0.6543192863464355, "learning_rate": 9.919063227033445e-05, "loss": 2.4174, "step": 7360 }, { "epoch": 0.06, "grad_norm": 0.5751597285270691, "learning_rate": 9.918952007752776e-05, "loss": 2.3064, "step": 7365 }, { "epoch": 0.06, "grad_norm": 1.7498908042907715, "learning_rate": 9.91884071273296e-05, "loss": 2.3316, "step": 7370 }, { "epoch": 0.06, "grad_norm": 0.7364647388458252, "learning_rate": 9.918729341975712e-05, "loss": 2.3122, "step": 7375 }, { "epoch": 0.06, "grad_norm": 1.0708686113357544, "learning_rate": 9.918617895482749e-05, "loss": 2.4102, "step": 7380 }, { "epoch": 0.06, "grad_norm": 0.7088509202003479, "learning_rate": 9.918506373255784e-05, "loss": 2.3707, "step": 7385 }, { "epoch": 0.06, "grad_norm": 0.7562710642814636, "learning_rate": 9.918394775296536e-05, "loss": 2.41, "step": 7390 }, { "epoch": 0.06, "grad_norm": 1.0215024948120117, "learning_rate": 9.91828310160672e-05, "loss": 2.3592, "step": 7395 }, { "epoch": 0.06, "grad_norm": 1.224595308303833, "learning_rate": 9.918171352188062e-05, "loss": 2.4744, "step": 7400 }, { "epoch": 0.06, "grad_norm": 5.069368362426758, "learning_rate": 9.918059527042275e-05, "loss": 2.3249, "step": 7405 }, { "epoch": 0.06, "grad_norm": 0.6148265600204468, "learning_rate": 9.917947626171086e-05, "loss": 2.3513, "step": 7410 }, { "epoch": 0.06, "grad_norm": 13.453545570373535, "learning_rate": 9.917835649576217e-05, "loss": 2.5281, "step": 7415 }, { "epoch": 0.06, "grad_norm": 0.9648706912994385, "learning_rate": 9.91772359725939e-05, "loss": 2.3651, "step": 7420 }, { "epoch": 0.06, "grad_norm": 1.2290962934494019, "learning_rate": 9.917611469222333e-05, "loss": 2.3721, "step": 7425 }, { "epoch": 0.06, "grad_norm": 0.7779213786125183, "learning_rate": 9.917499265466772e-05, "loss": 2.3461, "step": 7430 }, { "epoch": 0.06, "grad_norm": 74.53199005126953, "learning_rate": 9.917386985994432e-05, "loss": 2.5869, "step": 7435 }, { "epoch": 0.06, "grad_norm": 1.1777350902557373, "learning_rate": 9.917274630807047e-05, "loss": 2.3014, "step": 7440 }, { "epoch": 0.06, "grad_norm": 0.9880056381225586, "learning_rate": 9.917162199906341e-05, "loss": 2.3203, "step": 7445 }, { "epoch": 0.06, "grad_norm": 0.9910576343536377, "learning_rate": 9.917049693294049e-05, "loss": 2.4198, "step": 7450 }, { "epoch": 0.06, "grad_norm": 1.211416244506836, "learning_rate": 9.916937110971902e-05, "loss": 2.3295, "step": 7455 }, { "epoch": 0.06, "grad_norm": 0.7370124459266663, "learning_rate": 9.916824452941635e-05, "loss": 2.3977, "step": 7460 }, { "epoch": 0.06, "grad_norm": 1.339231014251709, "learning_rate": 9.916711719204981e-05, "loss": 2.2924, "step": 7465 }, { "epoch": 0.06, "grad_norm": 8.374075889587402, "learning_rate": 9.916598909763675e-05, "loss": 2.4362, "step": 7470 }, { "epoch": 0.06, "grad_norm": 1.3087061643600464, "learning_rate": 9.916486024619455e-05, "loss": 2.2994, "step": 7475 }, { "epoch": 0.06, "grad_norm": 0.7992890477180481, "learning_rate": 9.916373063774062e-05, "loss": 2.3551, "step": 7480 }, { "epoch": 0.06, "grad_norm": 3.2108137607574463, "learning_rate": 9.91626002722923e-05, "loss": 2.4482, "step": 7485 }, { "epoch": 0.06, "grad_norm": 0.5851492285728455, "learning_rate": 9.916146914986703e-05, "loss": 2.3284, "step": 7490 }, { "epoch": 0.06, "grad_norm": 1.7802926301956177, "learning_rate": 9.916033727048223e-05, "loss": 2.4944, "step": 7495 }, { "epoch": 0.06, "grad_norm": 2.11102557182312, "learning_rate": 9.915920463415531e-05, "loss": 2.4162, "step": 7500 }, { "epoch": 0.06, "grad_norm": 1.9146034717559814, "learning_rate": 9.915807124090371e-05, "loss": 2.3205, "step": 7505 }, { "epoch": 0.06, "grad_norm": 0.719231128692627, "learning_rate": 9.915693709074487e-05, "loss": 2.3494, "step": 7510 }, { "epoch": 0.06, "grad_norm": 3.0053324699401855, "learning_rate": 9.91558021836963e-05, "loss": 2.3657, "step": 7515 }, { "epoch": 0.06, "grad_norm": 0.6951493620872498, "learning_rate": 9.915466651977543e-05, "loss": 2.45, "step": 7520 }, { "epoch": 0.06, "grad_norm": 1.0601170063018799, "learning_rate": 9.915353009899976e-05, "loss": 2.2827, "step": 7525 }, { "epoch": 0.06, "grad_norm": 0.8246048092842102, "learning_rate": 9.915239292138678e-05, "loss": 2.2151, "step": 7530 }, { "epoch": 0.06, "grad_norm": 0.8652237057685852, "learning_rate": 9.915125498695402e-05, "loss": 2.3606, "step": 7535 }, { "epoch": 0.06, "grad_norm": 1.0338788032531738, "learning_rate": 9.9150116295719e-05, "loss": 2.3257, "step": 7540 }, { "epoch": 0.06, "grad_norm": 0.922222375869751, "learning_rate": 9.914897684769921e-05, "loss": 2.3663, "step": 7545 }, { "epoch": 0.06, "grad_norm": 0.9745861291885376, "learning_rate": 9.914783664291225e-05, "loss": 2.4035, "step": 7550 }, { "epoch": 0.06, "grad_norm": 2.2892563343048096, "learning_rate": 9.914669568137565e-05, "loss": 2.3931, "step": 7555 }, { "epoch": 0.06, "grad_norm": 0.8591157793998718, "learning_rate": 9.914555396310698e-05, "loss": 2.4133, "step": 7560 }, { "epoch": 0.06, "grad_norm": 2.102069616317749, "learning_rate": 9.914441148812382e-05, "loss": 2.3369, "step": 7565 }, { "epoch": 0.06, "grad_norm": 0.5565800070762634, "learning_rate": 9.914326825644376e-05, "loss": 2.4314, "step": 7570 }, { "epoch": 0.06, "grad_norm": 0.788996160030365, "learning_rate": 9.91421242680844e-05, "loss": 2.2649, "step": 7575 }, { "epoch": 0.06, "grad_norm": 0.5702605247497559, "learning_rate": 9.914097952306338e-05, "loss": 2.3989, "step": 7580 }, { "epoch": 0.06, "grad_norm": 0.6749225854873657, "learning_rate": 9.913983402139828e-05, "loss": 2.2833, "step": 7585 }, { "epoch": 0.06, "grad_norm": 0.5520764589309692, "learning_rate": 9.913868776310677e-05, "loss": 2.2819, "step": 7590 }, { "epoch": 0.06, "grad_norm": 0.7207875847816467, "learning_rate": 9.913754074820649e-05, "loss": 2.421, "step": 7595 }, { "epoch": 0.06, "grad_norm": 1.6550891399383545, "learning_rate": 9.913639297671511e-05, "loss": 2.2616, "step": 7600 }, { "epoch": 0.06, "grad_norm": 1.2345293760299683, "learning_rate": 9.91352444486503e-05, "loss": 2.4262, "step": 7605 }, { "epoch": 0.06, "grad_norm": 0.5169796943664551, "learning_rate": 9.913409516402974e-05, "loss": 2.4662, "step": 7610 }, { "epoch": 0.06, "grad_norm": 1.413092851638794, "learning_rate": 9.913294512287112e-05, "loss": 2.3154, "step": 7615 }, { "epoch": 0.06, "grad_norm": 0.7536594867706299, "learning_rate": 9.913179432519215e-05, "loss": 2.4231, "step": 7620 }, { "epoch": 0.06, "grad_norm": 1.0644789934158325, "learning_rate": 9.913064277101055e-05, "loss": 2.3841, "step": 7625 }, { "epoch": 0.06, "grad_norm": 1.828444242477417, "learning_rate": 9.912949046034406e-05, "loss": 2.3955, "step": 7630 }, { "epoch": 0.06, "grad_norm": 0.5712507963180542, "learning_rate": 9.912833739321042e-05, "loss": 2.324, "step": 7635 }, { "epoch": 0.06, "grad_norm": 1.149045467376709, "learning_rate": 9.912718356962738e-05, "loss": 2.4152, "step": 7640 }, { "epoch": 0.06, "grad_norm": 0.6611480712890625, "learning_rate": 9.91260289896127e-05, "loss": 2.3875, "step": 7645 }, { "epoch": 0.06, "grad_norm": 0.9991132616996765, "learning_rate": 9.912487365318417e-05, "loss": 2.4033, "step": 7650 }, { "epoch": 0.06, "grad_norm": 1.4412802457809448, "learning_rate": 9.912371756035957e-05, "loss": 2.4319, "step": 7655 }, { "epoch": 0.06, "grad_norm": 1.156378149986267, "learning_rate": 9.912256071115671e-05, "loss": 2.3287, "step": 7660 }, { "epoch": 0.06, "grad_norm": 0.9444717168807983, "learning_rate": 9.91214031055934e-05, "loss": 2.3549, "step": 7665 }, { "epoch": 0.06, "grad_norm": 0.5864918231964111, "learning_rate": 9.912024474368745e-05, "loss": 2.3641, "step": 7670 }, { "epoch": 0.06, "grad_norm": 1.003706693649292, "learning_rate": 9.911908562545671e-05, "loss": 2.3565, "step": 7675 }, { "epoch": 0.06, "grad_norm": 1.3801343441009521, "learning_rate": 9.911792575091903e-05, "loss": 2.4352, "step": 7680 }, { "epoch": 0.06, "grad_norm": 13.885682106018066, "learning_rate": 9.911676512009226e-05, "loss": 2.4027, "step": 7685 }, { "epoch": 0.06, "grad_norm": 1.5465410947799683, "learning_rate": 9.911560373299426e-05, "loss": 2.3826, "step": 7690 }, { "epoch": 0.06, "grad_norm": 0.8355954885482788, "learning_rate": 9.911444158964296e-05, "loss": 2.3973, "step": 7695 }, { "epoch": 0.06, "grad_norm": 0.6233800053596497, "learning_rate": 9.91132786900562e-05, "loss": 2.3723, "step": 7700 }, { "epoch": 0.06, "grad_norm": 0.5639955997467041, "learning_rate": 9.911211503425191e-05, "loss": 2.4804, "step": 7705 }, { "epoch": 0.06, "grad_norm": 0.5393581390380859, "learning_rate": 9.911095062224798e-05, "loss": 2.2716, "step": 7710 }, { "epoch": 0.06, "grad_norm": 3.9504611492156982, "learning_rate": 9.91097854540624e-05, "loss": 2.3526, "step": 7715 }, { "epoch": 0.06, "grad_norm": 0.8639929294586182, "learning_rate": 9.910861952971304e-05, "loss": 2.3596, "step": 7720 }, { "epoch": 0.06, "grad_norm": 0.6003659963607788, "learning_rate": 9.910745284921791e-05, "loss": 2.4116, "step": 7725 }, { "epoch": 0.06, "grad_norm": 1.4090920686721802, "learning_rate": 9.910628541259495e-05, "loss": 2.4028, "step": 7730 }, { "epoch": 0.06, "grad_norm": 4.262857437133789, "learning_rate": 9.91051172198621e-05, "loss": 2.3474, "step": 7735 }, { "epoch": 0.06, "grad_norm": 0.6540048718452454, "learning_rate": 9.910394827103742e-05, "loss": 2.3667, "step": 7740 }, { "epoch": 0.06, "grad_norm": 1.1263043880462646, "learning_rate": 9.910277856613886e-05, "loss": 2.3239, "step": 7745 }, { "epoch": 0.06, "grad_norm": 1.1847026348114014, "learning_rate": 9.910160810518442e-05, "loss": 2.4268, "step": 7750 }, { "epoch": 0.06, "grad_norm": 0.6711416840553284, "learning_rate": 9.910043688819218e-05, "loss": 2.3618, "step": 7755 }, { "epoch": 0.06, "grad_norm": 8.877666473388672, "learning_rate": 9.909926491518012e-05, "loss": 2.373, "step": 7760 }, { "epoch": 0.06, "grad_norm": 0.5303126573562622, "learning_rate": 9.909809218616629e-05, "loss": 2.4193, "step": 7765 }, { "epoch": 0.06, "grad_norm": 0.7092147469520569, "learning_rate": 9.909691870116877e-05, "loss": 2.3742, "step": 7770 }, { "epoch": 0.06, "grad_norm": 1.1718250513076782, "learning_rate": 9.909574446020561e-05, "loss": 2.3136, "step": 7775 }, { "epoch": 0.06, "grad_norm": 3.523742437362671, "learning_rate": 9.90945694632949e-05, "loss": 2.3762, "step": 7780 }, { "epoch": 0.06, "grad_norm": 1.265869140625, "learning_rate": 9.909339371045474e-05, "loss": 2.3894, "step": 7785 }, { "epoch": 0.06, "grad_norm": 0.7884125113487244, "learning_rate": 9.90922172017032e-05, "loss": 2.4649, "step": 7790 }, { "epoch": 0.06, "grad_norm": 4.220033168792725, "learning_rate": 9.909103993705845e-05, "loss": 2.4181, "step": 7795 }, { "epoch": 0.06, "grad_norm": 0.5695760250091553, "learning_rate": 9.908986191653857e-05, "loss": 2.3533, "step": 7800 }, { "epoch": 0.06, "grad_norm": 0.5501085519790649, "learning_rate": 9.90886831401617e-05, "loss": 2.3669, "step": 7805 }, { "epoch": 0.06, "grad_norm": 10.067580223083496, "learning_rate": 9.908750360794603e-05, "loss": 2.3771, "step": 7810 }, { "epoch": 0.06, "grad_norm": 0.866500973701477, "learning_rate": 9.908632331990968e-05, "loss": 2.3502, "step": 7815 }, { "epoch": 0.06, "grad_norm": 1.5686063766479492, "learning_rate": 9.908514227607084e-05, "loss": 2.3036, "step": 7820 }, { "epoch": 0.06, "grad_norm": 0.9078027606010437, "learning_rate": 9.90839604764477e-05, "loss": 2.4215, "step": 7825 }, { "epoch": 0.06, "grad_norm": 0.7654961347579956, "learning_rate": 9.908277792105845e-05, "loss": 2.4162, "step": 7830 }, { "epoch": 0.06, "grad_norm": 0.7100294232368469, "learning_rate": 9.90815946099213e-05, "loss": 2.2979, "step": 7835 }, { "epoch": 0.06, "grad_norm": 0.512353777885437, "learning_rate": 9.908041054305446e-05, "loss": 2.2994, "step": 7840 }, { "epoch": 0.06, "grad_norm": 1.215636968612671, "learning_rate": 9.907922572047618e-05, "loss": 2.3867, "step": 7845 }, { "epoch": 0.06, "grad_norm": 1.2812292575836182, "learning_rate": 9.90780401422047e-05, "loss": 2.4342, "step": 7850 }, { "epoch": 0.06, "grad_norm": 23.287656784057617, "learning_rate": 9.907685380825825e-05, "loss": 2.3487, "step": 7855 }, { "epoch": 0.06, "grad_norm": 1.1778185367584229, "learning_rate": 9.907566671865512e-05, "loss": 2.37, "step": 7860 }, { "epoch": 0.06, "grad_norm": 0.9992426037788391, "learning_rate": 9.907447887341359e-05, "loss": 2.4231, "step": 7865 }, { "epoch": 0.06, "grad_norm": 1.3622918128967285, "learning_rate": 9.907329027255194e-05, "loss": 2.3776, "step": 7870 }, { "epoch": 0.06, "grad_norm": 0.7109755277633667, "learning_rate": 9.907210091608848e-05, "loss": 2.3925, "step": 7875 }, { "epoch": 0.06, "grad_norm": 1.4260568618774414, "learning_rate": 9.907091080404152e-05, "loss": 2.3397, "step": 7880 }, { "epoch": 0.06, "grad_norm": 0.8834895491600037, "learning_rate": 9.906971993642937e-05, "loss": 2.413, "step": 7885 }, { "epoch": 0.06, "grad_norm": 7.023975849151611, "learning_rate": 9.906852831327038e-05, "loss": 2.301, "step": 7890 }, { "epoch": 0.06, "grad_norm": 1.36215078830719, "learning_rate": 9.906733593458289e-05, "loss": 2.325, "step": 7895 }, { "epoch": 0.06, "grad_norm": 0.7030614018440247, "learning_rate": 9.906614280038529e-05, "loss": 2.4227, "step": 7900 }, { "epoch": 0.06, "grad_norm": 0.6623185873031616, "learning_rate": 9.90649489106959e-05, "loss": 2.3465, "step": 7905 }, { "epoch": 0.06, "grad_norm": 0.49469253420829773, "learning_rate": 9.906375426553314e-05, "loss": 2.1638, "step": 7910 }, { "epoch": 0.06, "grad_norm": 0.626288890838623, "learning_rate": 9.90625588649154e-05, "loss": 2.4016, "step": 7915 }, { "epoch": 0.06, "grad_norm": 1.0985220670700073, "learning_rate": 9.906136270886107e-05, "loss": 2.4307, "step": 7920 }, { "epoch": 0.06, "grad_norm": 3.5122501850128174, "learning_rate": 9.90601657973886e-05, "loss": 2.4133, "step": 7925 }, { "epoch": 0.06, "grad_norm": 0.9121112823486328, "learning_rate": 9.905896813051636e-05, "loss": 2.2994, "step": 7930 }, { "epoch": 0.06, "grad_norm": 0.5800424814224243, "learning_rate": 9.905776970826286e-05, "loss": 2.3726, "step": 7935 }, { "epoch": 0.06, "grad_norm": 0.5281596183776855, "learning_rate": 9.905657053064651e-05, "loss": 2.2975, "step": 7940 }, { "epoch": 0.06, "grad_norm": 0.7939925789833069, "learning_rate": 9.905537059768578e-05, "loss": 2.3694, "step": 7945 }, { "epoch": 0.06, "grad_norm": 0.5826491713523865, "learning_rate": 9.905416990939915e-05, "loss": 2.3561, "step": 7950 }, { "epoch": 0.06, "grad_norm": 9.454414367675781, "learning_rate": 9.905296846580512e-05, "loss": 2.4724, "step": 7955 }, { "epoch": 0.06, "grad_norm": 0.6050140261650085, "learning_rate": 9.905176626692217e-05, "loss": 2.3676, "step": 7960 }, { "epoch": 0.06, "grad_norm": 0.6113778948783875, "learning_rate": 9.905056331276884e-05, "loss": 2.3695, "step": 7965 }, { "epoch": 0.06, "grad_norm": 0.5925760269165039, "learning_rate": 9.904935960336359e-05, "loss": 2.4083, "step": 7970 }, { "epoch": 0.06, "grad_norm": 0.788550615310669, "learning_rate": 9.904815513872503e-05, "loss": 2.3328, "step": 7975 }, { "epoch": 0.06, "grad_norm": 0.9591420888900757, "learning_rate": 9.904694991887164e-05, "loss": 2.4206, "step": 7980 }, { "epoch": 0.06, "grad_norm": 6.605620384216309, "learning_rate": 9.904574394382203e-05, "loss": 2.3112, "step": 7985 }, { "epoch": 0.06, "grad_norm": 0.6757184267044067, "learning_rate": 9.904453721359474e-05, "loss": 2.3695, "step": 7990 }, { "epoch": 0.06, "grad_norm": 3.4146552085876465, "learning_rate": 9.904332972820837e-05, "loss": 2.3317, "step": 7995 }, { "epoch": 0.06, "grad_norm": 0.7010883688926697, "learning_rate": 9.904212148768148e-05, "loss": 2.4222, "step": 8000 } ], "logging_steps": 5, "max_steps": 126689, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 1000, "total_flos": 1.970167211674829e+19, "train_batch_size": 16, "trial_name": null, "trial_params": null }